ATS-ERS Statement On Respiratory Muscle Testing

American Thoracic Society/European Respiratory Society
ATS/ERS Statement on Respiratory Muscle Testing

THIS JOINT STATEMENT OF THE AMERICAN THORACIC SOCIETY (ATS), AND THE EUROPEAN RESPIRATORY SOCIETY (ERS) WAS
ADOPTED BY THE ATS BOARD OF DIRECTORS, MARCH 2001 AND BY THE ERS EXECUTIVE COMMITTEE, JUNE 2001
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
1. Tests of Overall Respiratory Function
G. John Gibson, William Whitelaw, Nikolaos Siafakas
Static Lung Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dynamic Spirometry and Maximum Flow . . . . . . . . . . .
Maximum Voluntary Ventilation . . . . . . . . . . . . . . . . . . .
Arterial Blood Gases: Awake . . . . . . . . . . . . . . . . . . . . . .
Measurements during Sleep . . . . . . . . . . . . . . . . . . . . . . .
Tests of Respiratory Control . . . . . . . . . . . . . . . . . . . . . . .
Carbon Monoxide Transfer . . . . . . . . . . . . . . . . . . . . . . . .
Exercise Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5. Assessment of Respiratory Muscle Fatigue

Gerald S. Supinski, Jean Will Fitting, Franois Bellemare
521
521
522
522
523
524
525
526
526
2. Tests of Respiratory Muscle Strength

Malcolm Green, Jeremy Road, Gary C. Sieck, Thomas
Similowski
Pressure Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . .
Devices for Measuring Pressures . . . . . . . . . . . . . . . . . . .
Techniques for Pressure Measurement . . . . . . . . . . . . . .
Volitional Tests of Respiratory Muscle Strength . . . . . .
Pressures Obtained via Phrenic Nerve Stimulation . . . .
Abdominal Muscle Stimulation . . . . . . . . . . . . . . . . . . . .
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
528
528
530
531
535
542
542
3. Electrophysiologic Techniques for the Assessment of

Respiratory Muscle Function
Thomas K. Aldrich, Christer Sinderby, David K. McKenzie,
Marc Estenne, Simon C. Gandevia
Electromyography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Stimulation Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
548
554
556
557
4. Tests of Respiratory Muscle Endurance

Thomas Clanton, Peter M. Calverly, Bartolome R. Celli
Measures of Respiratory Muscle Activity
Used in Endurance Testing . . . . . . . . . . . . . . . . . . . . . . 559
Ventilatory Endurance Tests . . . . . . . . . . . . . . . . . . . . . . 562
Endurance to External Loads . . . . . . . . . . . . . . . . . . . . . . 564
Am J Respir Crit Care Med Vol 166. pp 518624, 2002

DOI: 10.1164/rccm.166.4.518
Internet address: www.atsjournals.org
Endurance of the Diaphragm . . . . . . . . . . . . . . . . . . . . . . 568

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
Types of Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571

Tests of Respiratory Muscle Fatigue . . . . . . . . . . . . . . . . 572
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578
6. Assessment of Chest Wall Function

Stephen H. Loring, Andre de Troyer, Alex E. Grassino
Pressures in the Chest Wall . . . . . . . . . . . . . . . . . . . . . . . . 580
Assessment of the Properties of the Relaxed
Human Chest Wall: Rahn Diagram . . . . . . . . . . . . . . . 580
Assessment of the Function of the Active
Chest Wall: Campbell Diagram . . . . . . . . . . . . . . . . . . . 581
Estimation of Ventilation Based on Chest
Wall Motion: Konno-Mead Diagram . . . . . . . . . . . . . . 582
Devices Used to Monitor Breathing:
Pneumograph, Magnetometer, and
Respiratory Inductive Plethysmograph . . . . . . . . . . . . 583
Optical Devices Used to Measure Chest
Wall Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
Inferring Respiratory Muscle Contribution to
Breathing from Chest Wall Motion . . . . . . . . . . . . . . . 584
Breathing from the EsophagealGastric
Pressure Relationship: Macklem Diagram . . . . . . . . . . 585
Breathing from PressureVolume Relationships . . . . 585
Inferring Diaphragm Activation and
Electromechanical Effectiveness from EMG . . . . . . . 585
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
7. Imaging Respiratory Muscle Function

Neil B. Pride, Joseph R. Rodarte
Transmission Radiography . . . . . . . . . . . . . . . . . . . . . . . . . 588
Ultrasound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Volumetric Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
Nuclear Medicine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
519
8. Tests of Upper Airway Function

Neil J. Douglas, Samuel T. Kuna
Electromyography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Upper Airway Resistance . . . . . . . . . . . . . . . . . . . . . . . . .
Indirect Laryngoscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fiberoptic Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Computed Tomographic Scanning . . . . . . . . . . . . . . . . . .
Magnetic Resonance Imaging . . . . . . . . . . . . . . . . . . . . . .
Acoustic Reflection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FlowVolume Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Polysomnography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Muscle Biopsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Strength, Fatigue, and Endurance of Upper
Airway Muscles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Site of Pharyngeal Airway Closure during Sleep . . . . . .
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9. Tests of Respiratory Muscle Function in Children

Claude Gaultier, Julian Allen, Sandra England
593
594
596
596
596
597
597
597
597
598
598
598
598
Physiology of the Developing Respiratory Pump . . . . . . 601

Tests of Respiratory Function . . . . . . . . . . . . . . . . . . . . . . 601
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
10. Assessment of Respiratory Muscle Function in the

Intensive Care Unit
Martin J. Tobin, Laurent Brochard, Andrea Rossi
Breathing Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
Lung Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611
Pressure Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . 611
Prediction of Weaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
Introduction
Over the last 25 years, great efforts have been made to develop techniques to assess respiratory muscle function. Research output in this area has progressively increased, with the
number of peer reviewed articles published on respiratory muscle function having increased remarkably during the 19952000
period compared with 19801985.
This official joint statement represents the work of an expert ATS/ERS committee, which reviewed the merits of currently known techniques available to evaluate respiratory muscle function. The statement consists of 10 sections, each addressing
a major aspect of muscle function or a particular field of application. Each section addresses the rationale for the techniques,

DOI: 10.1164/rccm.2102104
their scientific basis, the equipment required, and, when pertinent, provides values obtained in healthy subjects or in patients. Some of the techniques reviewed in this statement have
thus far been used primarily in clinical research and their full
potential has not yet been established; however, they are mentioned for the purpose of stimulating their further development.
Through continued efforts in the area of respiratory muscle
testing, it is anticipated that there will be further enhancement
of diagnostic and treatment capabilities in specialties such as
intensive care, sleep medicine, pediatrics, neurology, rehabilitation, sports medicine, speech therapy, and respiratory
medicine.
1. Tests of Overall Respiratory Function

Routine measurements of respiratory function, that is, volumes, flows, and indices of gas exchange, are nonspecific in relation to diagnosis but give useful indirect information about
respiratory muscle performance. On occasion, the presence of
respiratory muscle dysfunction is first suspected from the pattern of conventional respiratory function tests. More frequently, they are of use in assessing the severity, functional
consequences, and progress of patients with recognized muscle weakness.
STATIC LUNG VOLUMES

Rationale and Scientific Basis
The most frequently noted abnormality of lung volumes in patients with respiratory muscle weakness is a reduction in vital
capacity (VC). The pattern of abnormality of other subdivisions of lung volume is less consistent. Residual volume (RV)
is usually normal or increased, the latter particularly with
marked expiratory weakness (1). Consequently, total lung capacity (TLC) is less markedly reduced than VC, and the RV/
TLC and FRC/TLC ratios are often increased without necessarily implying airway obstruction.
The VC is limited by weakness of both the inspiratory muscles, preventing full inflation, and expiratory muscles, inhibiting full expiration. In addition to the direct effect of loss of
muscle force, reductions in compliance of both the lungs (2)
and chest wall (3) also contribute to the reduction of VC in patients with chronic respiratory muscle weakness. In severe
weakness, the TLC and VC relate more closely to lung compliance than to the distending force (4, 5) (Figure 1). The
mechanism of reduced lung compliance is unclear. Contrary to
earlier suggestions, it is probably not simply due to widespread microatelectasis (6). Static lung volumes may also be
affected in some patients by coexistent lung or airway disease.
Vital capacity, thus, reflects the combined effect of weakness
and the static mechanical load on the respiratory muscles.
In mild respiratory muscle weakness, VC is less sensitive
than maximum respiratory pressures. However, the curvilinear relation between VC and maximum inspiratory pressure
(5) (Figure 2) implies that, in more advanced disease, marked
reductions in VC can occur with relatively small changes in
maximum pressures.
In patients with isolated or disproportionate bilateral diaphragmatic weakness or paralysis, the VC shows a marked fall
in the supine compared with the erect posture because of the
action of gravitational forces on the abdominal contents. In
some patients, this postural fall may exceed 50%. In most normal subjects, VC in the supine position is 510% less than
when upright (7) and a fall of 30% or more is generally associated with severe diaphragmatic weakness (8).
Disadvantages
VC has poor specificity for the diagnosis of respiratory muscle

weakness. In mild weakness, it is generally less sensitive to
changes than are maximum pressures (13).
Applications
Serial measurements of VC should be routine in monitoring

progress of patients with acute and chronic respiratory muscle
weakness.
Measurement of postural change of VC gives a simple index of weakness of the diaphragm relative to the other inspiratory muscles.
DYNAMIC SPIROMETRY AND MAXIMUM FLOW

Methodology and Equipment
Airway resistance is normal in uncomplicated respiratory

muscle weakness (14). Airway function may appear to be supernormal when volume-corrected indices such as FEV1/VC
or specific airway conductance are used (2).
The maximum expiratory and maximum inspiratory flow
volume curves characteristically show a reduction in those flows
that are most effort dependent, that is, maximum expiratory
flow at large lung volumes (including peak expiratory flow) and
maximum inspiratory flow at all lung volumes (2, 5) (Figure 3).
The descending limb of the maximum expiratory flowvolume
curve may suggest supernormal expiratory flow when this is related to absolute volume (2, 3). With severe expiratory weakness, an abrupt fall in maximum expiratory flow is seen immediately before RV is reached (1). In health the FEV1 is usually
less than the forced inspiratory volume in 1 second. Reversal of
this ratio is seen with upper (extrathoracic) airway obstruction,
as well as in respiratory muscle weakness, and may give a
pointer to these diagnoses during routine testing.
The effect of coughing can be visualized on the maximum expiratory flowvolume curve in healthy subjects as a transient flow
exceeding the maximum achieved during forced expiration. The
absence of such supramaximal flow transients during coughing
presumably results in impaired clearance of airway secretions
and is associated with more severe expiratory muscle weakness
(15). Even with quadriplegia, however, some patients can generate an active positive pleural pressure in expiration (16). This can
allow them to achieve the pressure required for flow limitation
through most of expiration so that FEV1 may still be reliable as
an index of airway function. Impaired maximal flow in some
neuromuscular diseases may also reflect poor coordination of
the respiratory muscles rather than decreased force per se.
Oscillations of maximum expiratory and/or inspiratory
flowthe so-called sawtooth appearanceare seen particularly when the upper airway muscles are weak and in patients
with extrapyramidal disorders (17) (Figure 4).
Recommendations and requirements for the measurement of

VC and other lung volumes are covered in detail elsewhere (9, 10).
Advantages
VC has excellent standardization, high reproducibility and wellestablished reference values. It is easily performed, widely available, and economical. It is quite sensitive for assessing progress
in moderate to severe respiratory muscle weakness. The rate of
decline has been shown to predict survival in both amyotrophic
lateral sclerosis (11) and Duchenne muscular dystrophy (12).
Recommendations and requirements for maximum flowvolume curves are covered in detail elsewhere (9, 10).
Advantages
Maximum flowvolume curves are easily performed, widely

available, and economical. Peak expiratory flow can be obtained with simple portable devices.
522
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
VOL 166
2002
Figure 1. Relation between static

lung compliance and total lung
capacity in 25 patients with
chronic respiratory muscle weakness of varying severity. Dashed
line is regression line. Reprinted
by permission from Reference 5.
Disadvantages
Intersubject
variability is greater than for VC. Reference val
ues for VEmax at standard percentages of FVC may present
problems of interpretation.
Applications
Visual inspection may suggest the likelihood of weakness.

The sawtooth appearance in an appropriate context may
suggest weakness or dyscoordination of upper airway muscles.
However, this appearance is nonspecific and is seen also in
some subjects with obstructive sleep apnea, nonapneic snoring, and thermal injury of the upper airway.
MAXIMUM VOLUNTARY VENTILATION

The maximum voluntary ventilation was formerly recommended

as a more specific test for muscle weakness than volume measurements but, in practice, the proportionate reduction is usually similar to that of VC (18, 19). Disproportionate reductions
may be seen in Parkinsons disease (20), in which the ability to
perform frequent alternating movements is impaired.
Recommendations and requirements are covered elsewhere (10).

Advantages
No advantages are perceived in most situations.

Disadvantages
The test depends on motivation and is tiring for the subject.
Figure 3. Schematic maximum expiratory and inspiratory flowvolume curves in a patient with severe respiratory muscle weakness (solid
line) compared with predicted (dotted line). Volume is expressed in absolute
terms (i.e., percent predicted).
Note marked reductions in FVC,
VEmax at higher volumes, and VImax at all volumes. Note also the
blunted
contour of the expiratory curve and the abrupt cessation of
VEmax at RV. In the midvolume range, VEmax exceeds that predicted

for the absolute lung volume.
ARTERIAL BLOOD GASES: AWAKE

In chronic muscle weakness, even when quite severe, PaO2 and

the alveolararterial PO2 difference are usually only mildly abnormal (2, 21). In acute muscle weakness, PaO2 may be more
markedly reduced, but the picture may be complicated by
atelectasis or respiratory infection (22).
With mild weakness, PaCO2 is usually less than normal (19,
22), implying alveolar hyperventilation. In the absence of primary pulmonary disease, daytime hypercapnia is unlikely unless respiratory muscle strength is reduced to 40% of predicted and VC is reduced to 50% of predicted (19) (Figures
5 and 6). Elevation of venous bicarbonate concentration occasionally gives an important clue to otherwise unsuspected hypercapnia. Patients with muscle weakness are less able than
normal subjects to compensate for minor changes in respiratory function. If hypercapnia is established or incipient, even
minor infections may cause a further rise in PaCO2, as also may
injudicious use of sedative drugs or uncontrolled oxygen.
Applications
Advantages
Maximum voluntary ventilation is not generally recommended

for patients with known or suspected respiratory muscle weakness but may be helpful in the assessment and monitoring of
patients with extrapyramidal disorders.
Arterial blood gases assess the major functional consequence

of respiratory muscle weakness. In patients with Duchenne
muscular dystrophy, hypercapnia has been shown to predict
shorter survival (12).
Figure 2. Curvilinear relation of

maximum static inspiratory pressure (inspiratory muscle strength)
to vital capacity in 25 patients with
chronic weakness of varying severity. Dashed line and statistics relate
to logarithmic regression. Solid line
represents relationship calculated
from a standard maximal static pressurevolume diagram assuming normal elastic properties of the respiratory system. The greater than
expected reduction in VC is due
to reduced compliance of the lungs
and chest wall. Reprinted by permission from Reference 5.
Figure 4. Maximum expiratory and

inspiratory flowvolume curves, showing sawtooth oscillations of flow.
523

Disadvantages
Figure 6. Relation of daytime PaCO2

to VC in 37 patients with uncomplicated chronic myopathy (closed
circles, regression line) and 16 with
myopathy plus chronic lung disease
(open circles). Reprinted by permission from Reference 19.
Definitely abnormal arterial blood gases usually imply late and

severe impairment of respiratory muscles and therefore their
measurement is neither sensitive nor specific. Daytime values
may underestimate the severity of abnormal gas exchange.
Applications
Measurement of arterial blood gases is routinely performed to

assess the consequences of respiratory muscle weakness.
MEASUREMENTS DURING SLEEP

Patients with moderate or severe respiratory muscle weakness

characteristically show dips in oxygen saturation (SaO2) related
to periods of rapid eye movement (REM) sleep (23, 24) (Figure 7). The episodic desaturation is usually due to hypopnea
and less often to apnea and is associated particularly with phasic REM sleep, when brief periods of rapid, irregular eye
movements are accompanied by reduced activity of skeletal
muscles (24) (Figure 8). The hypopneas and/or apneas may
appear to be either central (Figure 8) or obstructive, or
sometimes a mixture of both. The precise pattern of such
events depends on the relative activation of the respiratory
pump and upper airway dilator muscles (24). Obstructive apneas are more likely in weak patients who are also overweight
(25). In patients with severe respiratory muscle weakness, some
apneas that appear to be central may in fact be obstructive, incorrect classification being due to failure of external sensors to
detect chest wall movements of reduced amplitude (26).
Hypercapnia in patients with slowly progressive weakness
probably develops first during sleep. Continuous monitoring
during sleep (e.g., with a transcutaneous PCO2 electrode) shows a
gradual rise in PCO2 during REM sleep (23) (Figure 7). Consequently, PaCO2 measured shortly after waking is more likely to be
elevated than values obtained later in the day. Symptoms of nocturnal hypoventilation include morning headaches, daytime
sleepiness, and lack of energy. Similar symptoms can also result
from sleep disruption associated with frequent apneas and hypopneas, even in the absence of persistent hypercapnia. Daytime
somnolence is particularly common in patients with myotonic
dystrophy. However, even though sleep hypopnea and apnea are
frequently found in this condition, they appear not to explain the
sleepiness of most patients with myotonic dystrophy (27).
The timescale of progression from nocturnal to persistent
diurnal hypercapnia in patients with chronic respiratory muscle weakness is not known.
cause of apneas or hypopneas may require use of a supraglottic or

esophageal pressure sensor. Interpretation of recordings obtained
by inductance plethysmography or other devices that measure rib
cage and abdominal expansion is problematic in patients with quadriplegic or diaphragm paralysis. It is essential to check the polarity
of the tracings and to compare phase relations awake and asleep.
Reliability of the devices for monitoring PCO2 in sleep is
currently doubtful and requires more study.
Advantages
Overnight oximetry is simple to perform.

Nocturnal measurements are more sensitive for detection
of abnormal pulmonary gas exchange than daytime blood gases.
Disadvantages
Polysomnography is labor-intensive and relatively expensive.

Current evidence suggests that nocturnal hypoxemia is a less
good prognostic indicator than either vital capacity or awake
PaCO2 (12, 29).
Applications
The role of sleep measurements in patients with respiratory

muscle weakness is currently uncertain. Polysomnography may
be useful in patients with daytime sleepiness and suspected
nocturnal hypoventilation, perhaps especially if awake PaCO2
is borderline or only mildly elevated.
Marked REM-related desaturation is seen occasionally in
patients with relatively normal daytime SaO2 (26). More typically, however, the severity of nocturnal desaturation is predictable from daytime measurements, with more marked desaturation in patients with lower daytime PaO2, higher PaCO2,
and lower VC (23) (Figure 9).
Methodology
Polysomnographic techniques are described in detail elsewhere

(28). To assess whether upper airway narrowing is a contributing
Figure 5. Relation of daytime PaCO2

to respiratory muscle strength
(RMS arithmetic mean of PImax
and PEmax) in 33 patients with
uncomplicated chronic myopathy (closed circles, regression lines)
and 14 patients with myopathy plus
chronic lung disease (open circles).
Note that in uncomplicated myopathy, PaCO2 is reduced ( 40
mm Hg) in most patients with
mild weakness and is likely to be
elevated only when RMS 40%
predicted. Reprinted by permission from Reference 19.
Figure 7. Section of sleep recording of SaO2 and transcutaneous PCO2

(TcCO2) in a patient with chronic myopathy, showing mild desaturation (SaO2 90%) in non-REM sleep and frequent periodic dips in SaO2
in REM sleep. The PCO2 shows progressive elevation during REM periods. Reprinted by permission from Reference 23.
524
Figure 8. Brief ( 2 minute) polysomnographic recording in REM sleep

in a patient
with chronic myopathy. The signals are as follows: SaO2,
airflow ( V), posteroanterior motion of rib cage (RCPA) and abdomen
(ABPA), electro-oculogram (EOG), and integrated surface electromyograms from inspiratory intercostals (EMGint) and diaphragm (EMGdi)
(the ECG is superimposed on EMG signals). AD, Periods of REM sleep.
During periods A and C, marked irregular eye movements (phasic
REM) are accompanied by reduced EMG activity and consequently reduced motion and flow with subsequent desaturation; rib cage and
abdominal motion remain in phase, indicating central hypopneas. During periods B and D, eye movements are relatively quiescent and EMG activity increases with consequent increased motion and flow and subsequent recovery of SaO2 (increasing SaO2 during period C reflects the
increased ventilation in period B). Reprinted by permission from Reference 24.
Sleep studies should be performed in all patients for whom

nocturnal ventilatory support is being considered. On occasion, the finding of frequent hypopneas and/or apneas that are
predominantly obstructive will suggest a trial of treatment with
nasal continuous positive airway pressure. More frequently,
however, in patients with respiratory muscle weakness, bilevel
pressure support or another method of noninvasive intermittent positive pressure ventilation will be the treatment of
choice. Because there is no evidence that treatment of abnormalities of gas exchange per se during sleep is beneficial, currently there is no indication for widespread application of
polysomnography in the absence of relevant symptoms.
TESTS OF RESPIRATORY CONTROL

The respiratory control system may be considered to have

three functional components: (1) sensory receptors that provide information about the status of the respiratory system
(only chemoreceptors that measure arterial PCO2, PO2, and pH
are usually considered or tested, but there are many other sensory inputs of importance); (2) the central integrating circuits;
and (3) the motor output to the respiratory muscles. The tests
available are stimulus response tests, in which a receptor is
stimulated and the motor output or a downstream mechanical
effect of motor output, is measured. It is important to recognize that these tests are generally unable to separate the three
functional components of the control system.
Minute ventilation and arterial PCO2 are maintained at normal levels even with quite marked weakness of the respiratory
muscles, implying that the control system compensates for the
weakness by driving the respiratory muscles harder than normal. The mechanism by which the control system identifies
muscle weakness and adjusts its motor output is unknown. The
increased motor output is difficult to appreciate because it succeeds in generating only normal pressures, volumes, and flows.
It is most readily apparent when accessory muscles or abdominal muscles are more active than normal during quiet breathing.
VOL 166
2002
If phasic contraction of scalenes, sternocleidomastoids, pectoral muscles, or abdominal muscles can be palpated, it is safe to
conclude that respiratory motor output is above normal.
When respiratory muscles are chronically severely weak
and arterial PCO2 begins to rise, two explanations are possible.
The muscles may be so weak that they cannot continually generate sufficient alveolar ventilation. Otherwise, an abnormality of the ventilatory control system may be allowing the PCO2
to rise even though the muscles themselves are quite capable
of keeping it normal. A gradual shift in the PCO2 set point of
the controller does seem to occur in some patients with muscle
disease, as it does in some cases of sleep apnea and chronic
obstructive pulmonary disease.
Laboratory tests of overall respiration that have been used
to try to assess the control system include inhalation of hypercapnic or hypoxic gas mixtures to stimulate chemoreceptors,
with measurements of ventilation or occlusion pressure to assess motor output, and sleep studies to monitor behavior of
the control system during sleep.
In patients with weak muscles, interpretation of slopes of
conventional ventilatory curves is clouded for several reasons.
The output of the controller is abnormally high when ventilation is normal. The controller may therefore be on the
nonlinear part of its normal response curve.
The high motor neuron output cannot be measured directly
and its mechanical effect (e.g., ventilation) is reduced in the
presence of weakness.
The response will become flat if ventilation nears the limit of
respiratory muscle endurance and that limit may be only a
short distance above resting ventilation.
Abnormal central control of respiration is well documented
in bulbar poliomyelitis and other conditions affecting the central nervous system, presumably because of direct involvement of medullary respiratory centers. It has been suggested
that certain muscle diseases are also associated with primary
abnormalities of central respiratory control; these conditions
include myotonic dystrophy, acid maltase deficiency, and other
congenital myopathies. Impaired ventilatory responses to CO2
and/or hypoxia have frequently been described, but in many
cases, respiratory muscle function was assessed inadequately.
In myotonic dystrophy it has been shown that the relations between hypercapnia and both maximum respiratory pressures
and VC are similar to those in nonmyotonic diseases (30).
Occlusion pressure is the pressure generated in the airway
(and by inference the pressure generated in the pleural space)
by contraction of inspiratory muscles when the airway has
been occluded at end expiration. It was introduced to separate
hypoventilation due to high pulmonary resistance or elastance
from hypoventilation due to a failure of the respiratory pump
apparatus (i.e., the muscles, passive components of the chest
wall, and the control system) (31, 32). Occlusion pressure amplitude does not directly assess either the degree of muscle
weakness or the degree of neuronal adjustment to the weakness. P0.1 is the pressure generated in the first 100 milliseconds
of inspiration against an occluded airway. Its timing is such
that it is not influenced by the conscious response to occlusion
and as an index of ventilatory drive it has the advantage over
ventilation of being independent of the mechanical properties
of the lung (31). It is, however, dependent on the contractile
state and function of the respiratory muscles and consequently
on the lung volume at which it is measured. For example, because of the lengthtension relationship of the muscles, a reduced value for a given neural output would be expected with
pulmonary hyperinflation and an elevated FRC. On the other
525
change in either arterial PO2 or PCO2. A plot of PO2 (or PCO2)

against ventilation (or, for PO2 response the algebraic constants describing a hyperbola) are compared with normal values. The induced change in blood gases may be continuous
(rebreathing methods) or a few discrete points (steady state
methods). Usually PCO2 is held constant while PO2 is changed
and vice versa. Standard methods are available for measuring
ventilatory responses during rebreathing (37, 38).
Steady state or quasi-steady state tests (39) are done simply
by having the subject inhale a prepared mixture of gases, usually for 5 minutes (40). Judgments about the safety of inducing
hypoxemia or acidosis are made clinically for individual patients. In chronically hypoxemic patients, transient responses
to inhalation of pure oxygen may be useful and are safe (40).
For the measurement of P0.1, it is essential to close the airway exactly at the point of zero flow. This is usually done by
separating the inspiratory and expiratory lines with one-way
valves and then closing the inspiratory line while the subject is
exhaling. Conscious subjects must be unable to anticipate occlusions, which must be done silently and unexpectedly. Obstruction can be simply performed by inflating a balloon within
the lumen of the inspired line or by closing a valve. A sensitive
transducer and timer are used to record pressure at 0.1 second.
Advantages
Figure 9. Relation of sleep hypoxemia to daytime blood gases and VC in

20 patients with chronic myopathy (regression line [solid line] 95%
confidence limits [dashed lines]). The abscissa in each panel shows the
nadir SaO2 in REM sleep. More severe REM desaturation occurs with
lower awake PaO2 (top panel), higher awake PaCO2 (middle panel), and
lower VC (lower panel). Reprinted by permission from Reference 23.
hand, if inspiration starts below equilibrium lung volume the

value of P0.1 recorded depends on relaxation of the expiratory
muscles.
Values of P0.1 are around 1 cm H2O in normal subjects at
rest, around 3 cm H2O in patients with stable chronic obstructive pulmonary disease, and may be 10 cm H2O or more in
acute respiratory failure due to chronic obstructive pulmonary
disease or acute respiratory distress syndrome. Such values reflect a high ventilatory drive consequent on a greatly increased
mechanical load. Some, although not all, studies have suggested that in patients with chronic obstructive pulmonary disease receiving ventilatory support values greater than 46 cm
H2O are associated with failure to wean (33).
In patients with weak muscles, resting P0.1 tends to be normal or slightly increased (34). In the model of acute respiratory muscle weakness provided by partial curarization of
healthy subjects, the slope of P0.1 response to CO2 is increased
even though the ventilatory response is reduced (35). However, in patients with chronic weakness the ventilatory and P0.1
slopes are both diminished (even though resting P0.1 is normal
or increased). Hence, a reduced response in such individuals
does not necessarily imply impaired ventilatory drive (30).
A completely flat ventilatory response may identify defective

chemoreceptor or brainstem function, but lesser abnormalities
are difficult to interpret.
Occlusion pressure (P0.1) is relatively easy to measure.
Marked discrepancies between occlusion pressure and minute
ventilation point to a lung disease causing substantial increase
in airway resistance or lung elastance. Usually, however, such a
problem is clinically evident and better evaluated by spirometry.
Disadvantages
Indices of ventilatory control have a wide normal range and

are subject to overinterpretation.
Occlusion pressures in general, and P0.1 in particular, are
difficult to interpret without additional measurements of mechanics and control events through the whole respiratory cycle, which are usually not available. P0.1 is a valid index of neural output only at FRC. Breath-to-breath scatter in the data
requires averaging of many breaths to obtain precise results.
The theoretical issues regarding measurement and interpretation have been reviewed (41).
Clinical Applications
These tests are seldom used in routine clinical assessment of

stable patients. In acute respiratory failure, mouth occlusion pressure during unstimulated breathing may be of value in assessing respiratory drive and the likelihood of successful weaning.
Occlusion pressure has no proven clinical value in respiratory muscle disease but may occasionally be helpful by pointing to an unsuspected mechanical problem.
If a patient is known to have a mixed problem of muscle
weakness and a lung disease (e.g., polymyositis plus interstitial
pulmonary fibrosis) and the response of the controller to CO2
or O2 is being studied, P0.1 can be measured in conjunction
with ventilation as the response and may be a more reliable
way of comparing the result with normal values.
CARBON MONOXIDE TRANSFER
For assessment of ventilatory responses to hypercapnia or hypoxia (36), the subject inhales a gas mixture that causes a
Single-breath CO diffusing capacity (transfer factor) (DLCO)

in patients with muscle weakness is usually normal or mildly
526
reduced. Reduction is due to inability to achieve full distension of the lungs at TLC and consequent failure to expose all
the alveolar surface to carbon monoxide. As with other extrapulmonary causes of lung volume restriction, the transfer coefficient (KCO) is often supernormal.
Advantages
The measurement is easily performed and well standardized.

Disadvantages
A reduced DLCO is a nonspecific finding (but if accompanied

by elevation of KCO it suggests extrapulmonary volume restriction). Any effects of respiratory muscle weakness on the
measurements are indirect.
VOL 166
2002
airway muscles are weak and also in patients with extrapyramidal disorders (e.g., Parkinsons disease).
8. PaO2 and PaCO2 are affected by muscle weakness. Mild weakness causes slight hypoxemia and hypocapnia; severe weakness causes hypercapnia, but only when strength is 40%
predicted. A raised bicarbonate level may suggest muscle
weakness.
9. Respiratory muscle weakness may cause desaturation and
hypercapnia during REM sleep.
10. CO transfer (DLCO) in patients with muscle weakness is
normal or mildly reduced but, as with other causes of extrapulmonary lung volume restriction, the transfer coefficient (KCO) is often raised.
References
The pattern of normal or mildly reduced DLCO and raised KCO

directs attention to extrapulmonary conditions, that is, respiratory muscle weakness, pleural disease or rib cage abnormalities. Otherwise, the main role of measurement of CO uptake
is in the recognition or exclusion of coexistent lung disease.
1. Kreitzer SM, Saunders NA, Tyler HR, Ingram RH. Respiratory muscle function in amyotrophic lateral sclerosis. Am Rev Respir Dis 1978;117: 437447.
2. Gibson GJ, Pride NB, Newsom Davis J, Loh C. Pulmonary mechanics in
patients with respiratory muscle weakness. Am Rev Respir Dis 1977;
115:389395.
3. Estenne M, Heilporn A, Delhez L, Yernault J-C, De Troyer A. Chest
wall stiffness in patients with chronic respiratory muscle weakness.
Am Rev Respir Dis 1983;128:10021007.
4. Gibson GJ, Pride NB. Lung mechanics in diaphragmatic paralysis. Am
Rev Respir Dis 1979;119:119120.
5. De Troyer A, Borenstein S, Cordier R. Analysis of lung volume restriction
in patients with respiratory muscle weakness. Thorax 1980;35:603610.
6. Estenne M, Gevenois PA, Kinnear W, Soudon P, Heilporn A, De
Troyer A. Lung volume restriction in patients with chronic respiratory
muscle weakness: the role of microatelectasis. Thorax 1993;48:698701.
7. Allen SM, Hunt B, Green M. Fall in vital capacity with weakness. Br J
Dis Chest 1985;79:267271.
8. Laroche CM, Carroll N, Moxham J, Green M. Clinical significance of severe isolated diaphragm weakness. Am Rev Respir Dis 1988;138:862866.
9. Quanjer PH. Standardised lung function testing. Eur Respir J 1993;
6(Suppl 16):3S102S.
10. American Thoracic Society. Standardisation of spirometry: 1987 update.
Am Rev Respir Dis 1987;136:12851298.
11. Fallat RJ, Jewitt B, Bass M, Kamm B, Norris FH. Spirometry in amyotrophic lateral sclerosis. Arch Neurol 1979;36:7480.
12. Phillips M, Smith PEM, Carroll N, Edwards RHT, Calverley PMA.
Does nocturnal oxygen desaturation predict survival in childhood onset muscular dystrophy? Thorax 1997;52:A18.
13. Black LF, Hyatt RE. Maximal static respiratory pressures in generalised
neuromuscular disease. Am Rev Respir Dis 1971;103:641650.
14. Wesseling G, Quaedvlieg FCM, Wouters EFM. Oscillatory mechanics of the
respiratory system in neuromuscular disease. Chest 1992;102:17521757.
15. Polkey MI, Lyall RA, Green M, Leigh PN, Moxham J. Expiratory muscle function in amyotrophic lateral sclerosis. Am J Respir Crit Care
Med 1998;158:734741.
16. Estenne M, van Muylem A, Gorini M, Kinnear W, Heilporn A, de
Troyer A. Effects of abdominal strapping on forced expiration in tetraplegic patients. Am J Respir Crit Care Med 1998;157:9598.
17. Vincken WG, Cosio MG. Flow oscillations on the flowvolume loop:
clinical and physiological implications. Eur Respir J 1989;2:543549.
18. Serisier DE, Mastaglia FL, Gibson GJ. Respiratory muscle function and
ventilatory control. I. In patients with motor neurone disease; II. In
patients with myotonic dystrophy. Q J Med 1982;51:205226.
19. Braun NMT, Arora NS, Rochester DF. Respiratory muscle and pulmonary function in polymyositis and other proximal myopathies. Thorax
1983;38:616623.
20. Tzelepis GE, McCool FD, Friedman JH, Hoppin FG. Respiratory muscle dysfunction in Parkinsons disease. Am Rev Respir Dis 1988;138:
266271.
21. Lane DJ, Hazleman B, Nichols PJR. Late onset respiratory failure in patients with previous poliomyelitis. Q J Med 1974;43:551568.
22. Harrison BDW, Collins JV, Brown KGE, Clark TJH. Respiratory failure in neuromuscular diseases. Thorax 1971;26:579584.
23. Bye PTP, Ellis ER, Issa FG, Donnelly PM, Sullivan CE. Respiratory
failure and sleep in neuromuscular disease. Thorax 1990;45:241247.
24. White JES, Drinnan MJ, Smithson AJ, Griffiths CJ, Gibson GJ. Respiratory muscle activity and oxygenation during sleep in patients with
muscle weakness. Eur Respir J 1995;8:807814.
EXERCISE TESTING
In many patients with muscle weakness, exercise is limited,
and therefore, maximum oxygen consumption is reduced because of weakness of the leg muscles rather than cardiorespiratory factors. The limited available data suggest that the relation of workload to oxygen consumption is normal, as also are
indices of submaximal exercise performance (42).
Advantages
Formal testing allows confirmation and quantification of exercise incapacity and may aid elucidation of its mechanism.
Disadvantages
Exercise is limited by weakness of nonrespiratory muscles in

many patients with neuromuscular disease. Exercise testing is
poorly standardized in this patient population.
Exercise testing may help determine the main factor(s) limiting exercise capacity, especially if related or coexistent cardiac
or pulmonary disease is present or suspected.
CONCLUSION
This Section of the Statement has explored the usefulness of
analyzing the results of pulmonary function tests to infer alterations in respiratory muscle function. Some such inferences
are as follows:
1. Respiratory muscle weakness reduces VC.
2. Expiratory muscle weakness can increase RV.
3. Reduction in chest wall and lung compliance, as a consequence
of muscle weakness, reduces lung volumes, notably VC.
4. A fall in VC in the supine position, compared with when
upright, suggests severe diaphragm weakness or paralysis.
5. With respiratory muscle weakness the maximal expiratory
and inspiratory flowvolume loops show a reduction in effort-dependent flows (peak flows) and a sharp fall in endexpiratory flow.
6. Reduced maximal flows in neuromuscular disease may reflect poor respiratory muscle coordination.
7. Maximum inspiratory and expiratory flowvolume curves
showing sawtooth oscillations are seen when the upper
527

25. Labanowski M, Schmidt-Nowara W, Guilleminault C. Sleep and neuromuscular disease: frequency of sleep-disordered breathing in a neuromuscular disease clinic population. Neurology 1996;47:11731180.
26. Smith PEM, Calverley PMA, Edwards RHT. Hypoxaemia during sleep
in Duchenne muscular dystrophy. Am Rev Respir Dis 1988;137:884
888.
27. Gilmartin JJ, Cooper BG, Griffiths DJ, Walls TJ, Veale D, Stone TN,
Osselton JW, Hudgson P, Gibson GJ. Breathing during sleep in patients with myotonic dystrophy and non-myotonic respiratory muscle
weakness. Q J Med 1991;78:2131.
28. American Thoracic Society. Indications and standards for cardio-pulmonary sleep studies. Am Rev Respir Dis 1989;139:559568.
29. Gay PC, Westbrook PR, Daube JR, Litchy WJ, Windebank AJ, Iverson
R. Effects of alterations in pulmonary function and sleep variables on
survival in patients with amyotrophic lateral sclerosis. Mayo Clin Proc
1991;66:686694.
30. Gibson GJ, Gilmartin JJ, Veale D, Walls TJ, Serisier DE. Respiratory
muscle function in neuromuscular disease. In: Jones NL, Killian KJ,
editors. Breathlessness. Hamilton, Canada: CME; 1992. p. 6671.
31. Whitelaw WA, Derenne JP, Milic-Emili J. Occlusion pressure as a measure of respiratory center output in conscious man. Respir Physiol
1975;23:181199.
32. Matthews AW, Howell JBL. The rate of isometric inspiratory pressure
development as a measure of responsiveness to carbon dioxide in
man. Clin Sci Mol Med 1975;49:5768.
33. Sassoon CSH, Te TT, Mahutte CK, Light RW. Airway occlusion pressure: an
34.
35.
36.
37.
38.
39.
40.
41.
42.
important indicator for successful weaning in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1987;135:107113.
Baydur A. Respiratory muscle strength and control of ventilation in patients with neuromuscular disease. Chest 1991;99:330338.
Holle RHO, Schoene RB, Pavlin EJ. Effect of respiratory muscle weakness
on P0.1 induced by partial curarization. J Appl Physiol 1984;57: 11501157.
Cherniack NS, Dempsey J, Fencl V, Fitzgerald RS, Lourenco RV, Rebuck AS, Rigg J, Severinghaus JW, Weil JW, Whitelaw WA, et al.
Workshop on assessment of respiratory control in humans. I. Methods
of measurement of ventilatory responses to hypoxia and hypercapnia.
Am Rev Respir Dis 1977;115:177181.
Read DJC. A clinical method for assessing the ventilatory response to
CO2. Australas Ann Med 1967;16:20.
Rebuck AS, Campbell EJM. A clinical method for assessing the ventilatory response to hypoxia. Am Rev Respir Dis 1974;109:345.
Weil JW, Byrne-Quinn E, Sodal JE, Filey GF, Grover RF. Acquired attenuation of chemoreceptor function in chronically hypoxic man at altitude. J Clin Invest 1971;50:186.
Cunningham DJC, Cormack RS, ORiordan JLH, Jukes MGM, Lloyd
BB. An arrangement for studying the respiratory effects in man of
various factors. Q J Exp Physiol 1957;42:294.
Whitelaw WA, Derenne J-P. Airway occlusion pressure. J Appl Physiol
1995;74:14751483.
Carroll JE, Hagberg JM, Brooks MH, Shumate JB. Bicycle ergometry
and gas exchange measurements in neuromuscular disease. Arch Neurol 1979;36:457461.
2. Tests of Respiratory Muscle Strength

PRESSURE MEASUREMENTS
Muscles have two functions: to develop force and to shorten.
In the respiratory system, force is usually estimated as pressure and shortening as lung volume change or displacement of
chest wall structures. Thus, quantitative characterization of the
respiratory muscles has usually relied on measurements of volumes, displacements, pressures, and the rates of change of these
variables with time.
Several important considerations have to be kept in mind:
1. Pressures at a given point are usually measured as a difference from barometric pressure.
2. Pressures measured at a point are taken to be representative of the pressure in that space. Differences in pressure at
different locations in normal subjects can arise from two
causes: gravity and shear stress (1). Gravity causes vertical
pressure gradients related to the density of the contents of
the space. In the thorax this gradient is 0.2 cm H2O cm1
height and is related to lung density. In the abdomen, this
gradient is nearly 1 cm H2O cm1 height. Pressure fluctuations are usually little affected by gravitational gradients.
Deformation of shape-stable organs can cause local variations in pressure, such as those that occur when the diaphragm displaces the liver during a large forceful diaphragmatic contraction (2). Pleural pressure may not be uniform
in patients with disordered lung architecture, particularly
emphysema. The schematic drawing in Figure 1 shows relationships between pressures and intervening respiratory
structures and equipment.
3. Pressure differences across structures are usually the relevant
pressures for characterizing those structures. Table 1 lists
pressures measured at a point and pressure differences across
structures, which are usually taken in a direction such that
positive pressure differences inflate the structure or lung.
4. A pressure difference between two points is always the
pressure difference across two or more structures or groups
of structures. For example, the pressure difference between
the pleural space and the body surface in a breathing person is both the trans-chest wall (transthoracic) and the
transpulmonary pressure.
The relationship between pressure and force is complex. For
example, thoracic geometry plays a major role in the efficiency of the conversion of force into pressure. The latter also
depends on the mechanical characteristics of the rib cage and
abdominal wall with which respiratory muscles interact: a
stiffer rib cage better resists distortion and therefore allows
more pressure to be produced by the diaphragm for a given
level of force (3). It follows that pressures should be viewed as
indices of global respiratory muscle output rather than as
direct measures of their contractile properties. Phonomyography could in future provide information related to force (4,
5) (see also sections on fatigue).
To test respiratory muscle properties, pressures can be
measured either during voluntary maneuvers (see subsequent
section) or during involuntary contractions, notably in response to phrenic nerve stimulation (see subsequent section).
In the former, the synergistic action of several inspiratory or
DOI: 10.1164/rccm.AT0202
For Abbreviations
page 547.
Internet
address: see
www.atsjournals.org
expiratory muscle groups is tested. In the latter, the pressure

developed is specific to the contracting muscle(s).
The purpose of this article is to describe the methodology
used to measure the various pressures for the assessment of
respiratory muscle strength.
DEVICES FOR MEASURING PRESSURES

A comprehensive review of the techniques for measurement
of pressures in respiratory physiology and of the associated
problems was presented by Milic-Emili (6) in 1984.
Pressure Transducers
As for most pressure measurements of respiratory events, a

frequency response flat up to 1015 Hz is adequate to measure
both dynamic and static pressures related to contractions of
respiratory muscles. The frequency response of a transducer
can be much altered by the characteristics of the systems attached to it, including balloons, tubing, and interconnecting
fittings (7) (see subsequent section). Thus, testing the response
characteristics of any transducer with the specific connectors
and fittings that are to be used to make the measurements of
pressure is highly recommended (7).
When differential pressure transducers are used, care must
be taken that their two sides have identical frequency responses. Calibration is best made with water manometers.
Electrical calibration is acceptable, but should be checked regularly with a water manometer.
The required range and sensitivity of the transducers depends on the test in question. Phrenic nerve stimulation in disease may develop pressures as low as a few centimeters of water, whereas maximal static maneuvers in healthy subjects can
be associated with positive and negative pressures exceeding
200 cm H2O. It may be possible to use a single type of transducer for all respiratory muscles tests, provided that it is sufficiently sensitive, with a resolution of approximately 0.5 cm
H2O and a range 200 cm H2O. Pressure differences between
two points can be measured directly with two catheters connected to a single differential pressure transducer.
Excellent pressure transducers, with such characteristics,
are commercially available, including devices based on a metal
membrane. More recently, other types of transducer that provide good results (e.g., piezoelectric transducers) have been
made available at lower cost.
Probes for Internal Pressures
Balloon catheter systems. The balloon catheter system is the

most widely used method for recording esophageal pressure
(Pes, Poes; see APPENDIX for a list of abbreviations) as a reflection of pleural pressure (Ppl), and gastric pressure (Pga) as a
reflection of abdominal pressure (Pab) (8). Air-containing latex balloons are sealed over catheters, which in turn transmit
pressures to the transducers. Single- and double-balloon catheter systems are commercially available, but can be made inhouse at low cost. Double-balloon catheters associated with
an electromyograph (EMG) electrode have been used (911).
When choosing or preparing a balloon catheter system, careful attention must be given to its physical characteristics. Indeed, the volume of the balloon, its volumepressure characteristics, and the dimensions of the catheter can influence the
measurement of pressure and introduce major errors. Standardization has been proposed (12).
529

TABLE 1. PRESSURES FOR BASIC RESPIRATORY MECHANICS
Pressures at a location
Pao airway opening pressure
Palv alveolar pressure
Ppl pleural pressure
Pab abdominal pressure
Pbs body surface pressure
Pressure differences across structures
Pel(L) elastic recoil pressure of the Lung (pressure across lung tissue)
PL transpuLmonary pressure (also Ptp)
Prc pressure across the rib cage
Paw flow-resistive pressure in airways
Pcw pressure across the chest wall
Pdi transdiaphragmatic pressure
Prs transrespiratory system pressure
Pabw transabdominal wall pressure
Peq pressure across the equipment
Relationships among pressures
Paw Pao Palv

PL Pao Ppl
Pel(L) Palv Ppl
Prc Ppl Pbs

Pdi Ppl Pab
Pcw Ppl Pbs

Pabw Pab Pbs
Figure 1. Locations at which pressures can be measured, and pressure

differences derived from them (see also Table 1). AbW abdominal
wall; aw airway; Di diaphragm; Eq equipment; Lt lung tissue;
Pab abdominal pressure; Palv alveolar pressure; Pao pressure at
airway opening; Pbs body surface pressure; Ppl pleural pressure;
rc rib cage.
For the measurement of Pes, good results have been provided by latex balloons 510 cm long, 3.55 cm in perimeter, and
with a thin wall (8, 13, 14). For accurate transmission of pressure, air should be introduced into the balloon until it is fully
distended to smooth out folds, and then most of the air removed
so that a volume is retained at which the rubber is unstretched
without distending the esophagus significantly. A volume of 0.5
ml is adequate for balloons with these characteristics. The volume displacement coefficient of the balloon cathetertransducer system should be measured, particularly if the balloon will
measure positive pressures, to ensure that the pressure level to
be measured does not completely empty the balloon into the
catheter and transducer. Thus, if high positive pressures are to
be measured (e.g., for Pes during maximal expiratory maneuvers) a volume of 0.5 ml may be inadequate (6). Balloon volumes should be checked repeatedly during measurements.
For the measurement of Pga, balloon volume is less crucial
and measurements can be made with a balloon volume of 12
ml, given that this remains within the range of volume over
which the rubber is unstretched. If studies of relatively long
duration are planned, the walls of the gastric balloon should
be thicker than those of esophageal balloons to increase resilience to gastric secretions.
Respiratory muscle studies can involve dynamic maneuvers
with high rates of change in pressure (e.g., sniffs and twitches)
resulting in a significant risk of a damped signal if the frequency response of the measuring system is inadequate, as may
occur if the internal diameter of the catheter is too small or
the gas volume too large. Polyethylene catheters with an internal diameter 1.41.7 mm and 70100 cm in length provide,
when associated with adequate transducers, an appropriate
frequency response (6).
Prs Pao Pbs Peq
The catheter should be reasonably stiff, with a series of

holes arranged in a spiral pattern over the entire portion of
the catheter covered by the balloon, because the gas in the
balloon tends to shift to the point where the pressure surrounding it is most negative, i.e., the top of the balloon in upright subjects.
Liquid-filled catheters. Fluid-filled catheter systems have
been employed, mainly in neonates and small animals for
study of respiratory mechanics. Their advantage is that the
transmission of pressure involving a noncompressible fluid
(usually water) gives a high-frequency response. The catheters
can, thus, be thinner than for balloons, theoretically reducing
discomfort. An important practical difficulty is the need for
regular flushing of the catheter, to avoid plugging of distal
holes and to keep the cathetermanometer system free of air
bubbles, which may dampen the measured pressure. Another
drawback is that while the gas bubble in the balloon migrates
to the point where the pressure is least (which is thought to
minimize artifacts in the esophagus and to locate pressure at
the surface of the gastric air bubble in the stomach) in a liquidfilled catheter, pressure is always measured at the end of the
catheter, which may not be the optimal site. Respiratory muscle studies in adult humans with this technique are limited or
not described, and its place in this context is probably limited.
Catheter-mounted microtransducers. Catheter-mounted microtransducers, often referred to as Millar catheters (15, 16),
have a level of performance comparable to that of balloon catheters (17, 18). Their management during long studies is probably easier, with a lower risk of technical problems (e.g., leaking
balloons), and they may be easier to tolerate for the subject.
Their frequency response is high, which may eliminate the phase
lag sometimes seen with balloon catheters during extremely
rapid pressure changes. However, catheter-mounted microtransducers record pressure at a single focused point so that the
measured Pes may not be as representative of Ppl as balloon
catheters, which sample pressure at the point where it is most
negative. They are also much more expensive than balloon catheter systems, and may be difficult to sterilize and reuse with
confidence.
Other systems. Other systems exist to measure pressures in
humans, including fiberoptic sensors. Fiberoptic sensors have
long been used for measurement of intracerebral pressures in
530
neurosurgery (19) (for review, see Yellowlees [20] and Shapiro

and coworkers [21]). They are probably adequate to measure
respiratory pressures (22), and may offer advantages over other
devices, including decreased chance of false measurements
due to occlusion with water or mucus, less chance of kinking,
and, possibly, more rapid response to pressure changes. This
remains to be precisely established, and, apparently, no study
of fiberoptic systems in respiratory muscle tests is available.
Devices for Measurement of Airway Opening Pressure
Air-filled catheter systems are commonly used to measure

pressures in airways and at the mouth. Airway opening pressure (Pao) is usually sampled from a side tap (lateral pressure)
in a mouthpiece (Pmo), tracheal tube (Ptr), face mask (Pmask),
or from a nostril plug (Pnas) (23). For nasal pressure to reflect
airway pressure there must be free communication between
the nostrils and mouth, with nasal flows. If Pao is measured
from a side tap of a mouthpiece or a tracheal tube during a
maneuver that involves gas flow, the cross-section of the device through which the subject breathes must be large enough
to avoid measurement errors due to the Bernoulli effect (24).
In some cases, Pao serves to estimate alveolar pressure (PA,
Palv) during dynamic respiratory efforts made against an obstructed airway (e.g., mouth pressure response to phrenic
nerve stimulation). For Pao to reflect PA accurately the transmission of pressure from the alveoli to the airway has to be
very fast. The time constant of transmission is the product of
the flow resistance offered by the airways (Raw) and the compliance of the extrathoracic airways (Cuaw) including the mouth,
cheeks, and equipment. In practice the internal volume of the
measuring equipment (mouthpiece, face mask, tracheal tube)
contributes negligibly to the time constant (6), but should be
minimized in patients with an already increased time constant,
such as patients with chronic obstructive pulmonary disease
(COPD). The compliance of the cheeks can be minimized by
holding them rigid with the hands.
TECHNIQUES FOR PRESSURE MEASUREMENT

Esophageal, Gastric, and Transdiaphragmatic Pressures
Scientific basis. Transdiaphragmatic pressure (Pdi) is defined

as the difference between Ppl and Pab (13) and, in practice, is
generally equated to the difference between Pes and Pga, so
that Pdi Pga Pes (where Pes is usually, but not always,
negative). This is contrary to most pressures across a structure, which are taken at a direction such that positive pressures inflate (e.g., positive transpulmonary pressures inflate
the lung). For this reason Pdi is also sometimes defined as Pdi
Pes Pga. As the diaphragm is the only muscle in which contraction simultaneously lowers Pes and increases Pga, an increase in Pdi is, in principle, the result of diaphragmatic contraction unless there is passive stretching. An inspiratory effort
produced with a completely passive unstretched diaphragm is
associated with a negative change in Pes and Pga but no change
in Pdi. This assumes that changes in Pes or Pga induced by
mechanisms other than diaphragm contraction are uniformly
transmitted across the diaphragm from one compartment to
the other. This is probably true when the diaphragm is relaxed
(6, 13) at functional residual capacity (FRC), but may be modified when the diaphragm is stretched, as at low lung volumes.
Methodology. Pes and Pga are most often measured by
passing a pair of probes, generally balloon catheters (see previous passages), through the nose, following local anesthesia
of the nasal mucosa and pharynx. Their position is usually assessed by asking the subject to perform sharp sniff maneuvers
while monitoring the signal on an oscilloscope or computer
VOL 166
2002
screen. A simple technique is to advance both probes well into

the stomach, as judged by a positive deflection during a sniff
and then to withdraw one of them until the sniff-related pressure deflection first becomes negative, indicating that the balloon has entered the esophagus. It is then withdrawn a further
10 cm. The validity of the Pes measurement can be checked by
matching Pes to Pao during static Mueller (inspiratory) maneuvers (the dynamic occlusion test) (6, 12, 14). Displacement
of balloons is minimized by taping the catheters to the nose.
The distance between the nostril and the tip of the balloons
varies with the size of the subject, but is usually 3540 cm for
Pes and 5060 cm for Pga in adults.
Placing the probes becomes more difficult when the subject
cannot perform voluntary inspiration (e.g., with anesthetized
patients, diaphragmatic paralysis, cognitive impairment, or
muscle incoordination). The pressure signals during a swallow
can then be useful: A balloon is positioned in the esophagus if
swallowing is associated with a slow, powerful rise in pressure,
whereas if this does not occur the balloon is likely to be in the
stomach. Measurement of balloon distance from the nostril
can be a useful indication of its position.
It is advisable to measure Pes and Pga separately by using
two pressure transducers, with Pdi derived from a third differential pressure transducer or reconstructed electronically offline. This allows the investigator to monitor balloon position
and detect confounding events such as esophageal spasms, as
well as recording the three pressures independently. Resting
Pga is usually positive with respect to atmosphere due to hydrostatic pressure in the abdomen. For respiratory muscle measurements Pga is conventionally taken as zero at resting end
expiration.
Advantages. Pdi is specific for diaphragm contraction (see
previous passages). Separate measurements of Pes and Pga
provide information on the components of this contraction
and Pes on the inspiratory driving pressure (Pes/Pdi ratio).
Disadvantages. The procedures require the subjects cooperation and occasionally untrained healthy volunteers can
fail to increase Pdi because of lack of coordination, in the absence of any diaphragmatic abnormality (25). This is, however, unusual during the inspiratory phase of quiet breathing
at rest. The measurements are mildly uncomfortable, both initially (when swallowing the catheters) and during studies.
However, the discomfort of swallowing a thin catheter is small
compared with other established medical procedures and
scarcely invasive. Good-quality equipment and adequate
practice minimize the discomfort, but some skill is necessary
and passing the probes can be time-consuming. Particular care
must be taken in patients with impaired swallowing, as well as
esophageal diseases, or disorders at the level of the gastroesophageal sphincter.
Mouth Pressure and Nostril Pressure
Scientific basis. Pmo is easy to measure and changes may give

a reasonable approximation of change in alveolar pressure
and thus Pes, providing there is relatively little pressure loss
down the airways, or across the lungs. This may be realistic
with normal lungs, particularly when changes in lung volume
are small, but is unlikely to be fulfilled in patients with severe
lung or airway disease. When used in combination with voluntary static and dynamic maneuvers at FRC, Pmo provides a
global index of the action of synergistic respiratory muscles.
When the diaphragm contracts in isolation against a closed
airway, as with phrenic nerve stimulation, Pmo may be a useful reflection of Pdi.
Pnas is also easy to measure (see VOLITIONAL TESTS OF RESPIRATORY MUSCLE STRENGTH) but has the same caveats as Pmo.
Methodology. Pmo is measured at the side port of a

mouthpiece. It should be possible to occlude the mouthpiece
at the distal end and a small leak should be incorporated to
prevent glottic closure during inspiratory or expiratory maneuvers (26). The type of mouthpiece used can significantly
influence the results (27). The issue of the lung volume at
which Pmo should be measured during static efforts is addressed in the section on volitional tests (see subsequent section), and the various maneuvers that can be used to obtain
useful Pmo data during phrenic nerve stimulation are described in the section on phrenic nerve stimulation (see subsequent section).
Pnas is measured with a polyethylene catheter held in one
nostril by a soft, hand-fashioned occluding plug; respiratory
maneuvers are performed through the contralateral nostril (23).
A standard mouthpiece for Pmo, or a nasal plug (custom
made or commercially available) for Pnas, and one pressure
transducer are required. Portable Pmo devices (28) are useful
for screening and bedside studies.
Advantages of mouth pressure and nasal sniff pressure.
The main advantage of Pmo and Pnas are their simplicity and
ease of use, both for the operator and for the subject.
Disadvantages of mouth pressure and nasal sniff pressure.
The measurement of Pmo does not allow the investigator to
discriminate between weakness of the different respiratory
muscles. When Pmo or Pnas is used as a substitute for Pes during dynamic maneuvers (sniff test, phrenic nerve stimulation),
glottic closure or airway characteristics may prevent adequate
equilibration.
VOLITIONAL TESTS OF RESPIRATORY

MUSCLE STRENGTH
The principal advantage of volitional tests is that they give an
estimate of inspiratory or expiratory muscle strength, are simple to perform, and are well tolerated by patients. Passage of
balloon catheter systems into the esophagus and/or stomach is
not usually required. However, it can be difficult to ensure
that the subject is making a truly maximal effort. Although
normal subjects can potentially activate peripheral and respiratory muscles fully during voluntary efforts (29), even experienced physiologists cannot always do this reliably for respiratory
efforts (30) and naive subjects have even greater difficulty
(31). Thus, it is hard to be certain whether low mouth pressure
measurements truly represent reduced strength, or merely reduced neural activation. Indeed, there may be some activation
of agonist muscles simultaneously (32). However, in practice a
normal result can be of value in precluding clinical weakness.
Maximal Static Inspiratory and Expiratory Pressure
Scientific basis. Measurement of the maximum static inspiratory pressure that a subject can generate at the mouth (PImax)
or the maximum static expiratory pressure (PEmax) is a simple
way to gauge inspiratory and expiratory muscle strength. The
pressure measured during these maneuvers reflects the pressure developed by the respiratory muscles (Pmus), plus the
passive elastic recoil pressure of the respiratory system including the lung and chest wall (Prs) (Figure 2 [33]). At FRC, Prs
is zero so that Pmo represents Pmus. However, at residual volume (RV), where PImax is usually measured, Prs may be as
much as 30 cm H2O, and thus makes a significant contribution to PImax of up to 30% (or more if Pmus is decreased).
Similarly, PEmax is measured at total lung capacity (TLC),
where Prs can be up to 40 cm H2O. Clinical measures and
normal values of PImax and PEmax do not conventionally subtract the respiratory system recoil.
531
The mouth pressures recorded during these maneuvers are

assumed to reflect respiratory muscle strength (Pmus) if Prs is
subtracted. However, maximum muscle strength in skeletal
muscles is the force developed under isometric conditions with
a muscle at its optimal length. In generating pressures during
respiratory maneuvers, muscle shortening (or lengthening) may
occur, with changes in forcevelocity and forcelength relationships (3436). The relationship between the tension (force) generated by a respiratory muscle (strength) and the pressure
produced in the thorax or mouth is complex. The diaphragm is
both a curved structure and acts as a piston so that the pressure or force per unit area output is only indirectly related to
muscle tension. In addition, the mechanical linkage of each individual respiratory muscle within the chest wall and with
other inspiratory or expiratory muscles influences the net
pressure produced. Thus, even though activation may be maximal, the pressure produced is derived from a complex set of
interactions within and between muscles and the chest wall
and its contents. Nevertheless, it is the pressure developed by
the inspiratory muscles that drives ventilation and, in spite of
the many assumptions, these measures can usefully reflect global respiratory muscle strength for clinical evaluation as well
as physiological studies. Thus, when respiratory muscle weakness occurs, the PImax can be more sensitive than the VC because the relationship between VC and PImax is curvilinear
(37), so that decreases in respiratory muscle strength occur before decreases in lung volume can be identified. On the other
hand, between- and within-individual variation in muscle
strength is considerably greater than that for vital capacity.
Between-individual variability may reflect the large variations
in strength in normal individuals.
Because of the forcelength relationship and the varying
contribution of Prs, PImax and PEmax vary markedly with lung
volume (38). Subjects find it easier to maximize their inspiratory efforts at low lung volumes and expiratory efforts at high
volumes; therefore, by convention and to standardize measurement, PImax is measured at or close to RV and PEmax at
or close to TLC. In some laboratories PImax and PEmax are
measured at FRC, and this may be more accurate for certain
research studies, but in this case the lung volume should be
specifically stated (39). In patients with abnormally high lung
volumes (e.g., patients with COPD), a low PImax may partly
reflect the shortened inspiratory muscle fiber length associated with increased lung volume at RV rather than reduced
inspiratory muscle strength (Figure 3). Furthermore, hyperinflation is often associated with intrinsic positive end-expiratory
pressure (PEEPi), so inspiratory efforts start from a negative
airway pressure. Thus, if PImax is measured as the maximal
negative airway pressure, it will underestimate the actual pressure generated by the inspiratory muscles. Optimally, under
such circumstances, PImax should be measured as the total
negative deflection of the occluded airway pressure during the
inspiratory effort, including the effort required to draw down
PEEPi.
Methodology. A number of authors have reported normal
values for PImax and PEmax (see Table 2 [26, 4044]). The
variation between these results presumably indicates differences between the groups studied and the way in which the
tests were performed and measured. Here, we propose a standardized approach to test performance and measurement.
Flanged mouthpieces are readily available in pulmonary
function laboratories and although they give values somewhat
lower than those obtained with a rubber tube mouthpiece, the
differences are not usually material in a clinical setting (27).
These mouthpieces are also easier for patients to use, especially
those with neuromuscular weakness. The flanged mouthpiece
532
Figure 2. Relationship of muscle and respiratory pressures at different

lung volumes. Vertical axis: lung volume as a percentage of vital capacity (%VC). Horizontal axis: alveolar pressure in cm H2O. The broken
lines indicate the pressure contributed by the muscles. Pmus pressure developed by the respiratory muscles; Prs pressure of the respiratory system. Reprinted by permission from Reference 33.
can be attached to a short, rigid tube with a three-way tap or

valve system to allow normal breathing followed by either a
maximum inspiratory or expiratory maneuver (Figure 4). For
research studies it may be preferable to use a rubber tube as
mouthpiece (26). However, this has to be held tightly around
the lips, to prevent leaks. This can be difficult for patients and
naive subjects particularly at high pressures, leading to significant pressure losses. The system requires a small leak (approximately 2-mm internal diameter [id] and 2030 mm in length)
to prevent glottic closure during the PImax maneuver and to
reduce the use of buccal muscles during the PEmax maneuver.
The inspiratory and expiratory pressure must be maintained,
ideally for at least 1.5 seconds, so that the maximum pressure
sustained for 1 second can be recorded. The peak pressure
may be higher than the 1 second of sustained pressure but is
believed to be less reproducible.
Historically, the aneroid manometer was used to measure the
pressure but this is not recommended as the analog signal on
the dial can be difficult to read accurately and pressure transients are difficult to eliminate. Mercury should be avoided for
safety reasons. A recording system should be used to collect
the pressure data and display it in analog form (strip chart recorder), or it can be digitized and displayed for measurement
(28) or the 1-second average computed (Figure 5). The pressure
transducers should be calibrated regularly against a fluid manometer with baseline pressure equal to atmospheric pressure.
The test should be performed by an experienced operator,
who should strongly urge subjects to make maximum inspiratory (Mueller maneuver) and expiratory (Valsalva maneuver)
efforts at or near RV and TLC, respectively. Subjects are normally seated and noseclips are not required. Because this is an
unfamiliar maneuver, careful instruction and encouraged motivation are essential. Subjects often need coaching to prevent
air leaks around the mouthpiece and to support the cheeks
during the expiratory efforts, and this may be helped by having them pinch their lips around the mouthpiece. Once the operator is satisfied, the maximum value of three maneuvers that
vary by less than 20% is recorded. Less variability may be necessary in a research setting, but even low variability may not
guarantee that maximal efforts have been made (45).
Advantages. The pressures measured at the mouth during
maximum inspiratory or expiratory maneuvers are widely
used specific tests of respiratory muscle strength. Normal values are available for adults, children, and the elderly. The tests
are not complicated to perform and are well tolerated by patients. The recent development of hand-held pressure meters
means the technique may be easily used at the bedside (28).
VOL 166
2002
Figure 3. Relationship between maximal static respiratory pressure

(PImax, PEmax) and lung volume. Pressures are expressed as a percentage of maximum and the lung volume is expressed as a percentage of
TLC. Symbols are data from different studies (39). Example A represents a patient with emphysema whose RV is 85% TLC, at which lung
volume his predicted PImax is only 50% of that at normal RV. Conversly, Example B represents a patient with lung fibrosis with TLC of
55% predicted, at which volume her P Emax is 82% maximum. Reprinted by permission from Reference 39.
Disadvantages. These tests are volitional and require full

cooperation. Accordingly, a low result may be due to lack of
motivation and does not necessarily indicate reduced inspiratory or expiratory muscle strength.
Normal values and applications. The recorded values of PImax
and PEmax may be compared with published normal values
(Table 2). The values that most closely reflect the protocol described here with a flanged mouthpiece, are those obtained by
Wilson and coworkers (43). Normal values for the elderly (46
48) and children (43, 4951) have been reported. The normal
ranges are wide (Table 2), so that values in the lower quarter
of the normal range are compatible both with normal strength
and with mild or moderate weakness. However, a PImax of
80 cm H2O usually excludes clinically important inspiratory
muscle weakness. Values less negative than this are difficult to
interpret and in such circumstances it would be appropriate to
undertake more detailed studies. A normal PEmax with a low
PImax suggests isolated diaphragmatic weakness.
Regional measurements. Static respiratory muscle pressures
generated against a closed airway can be recorded from balTABLE 2. REFERENCE NORMAL RANGES FOR PEmax AND PImax*
No.
PEmax
PImax
Source (Ref.)
Mouthpiece Design
23.4 4.5
22.8 4.1
21.2 4.4
15.1 8.0
14.4 3.3
13.7 3.7
12.7 3.1
12.1 2.1
12.4 2.7
11.1 3.5
10.4 3.0
10.3 2.5
40
26
41
42
43
44
Tube
Tube
Tube
Flanged
Flanged
Flanged
16.1 2.9
14.9 2.6
13.5 6.7
9.2 3.2
9.1 1.6
8.7 2.3
9.6 2.4
8.5 1.5
8.9 2.4
7.0 2.6
7.2 2.1
6.9 2.3
40
26
41
42
43
44
Tube
Tube
Tube
Flanged
Flanged
Flanged
Male
106
60
80
325
80
46
Female
94
60
121
480
87
60
Definition of abbreviations: PEmax maximum static expiratory pressure; PImax

maximum static inspiratory pressure.
* Values represent kilopascals (1 kPa 10.19 cm H2O), mean SD.
Reprinted by permission from Reference 27.
533
Advantages and disadvantages for regional measurements.

The measurement of maximum static transdiaphragmatic pressure, PI,di,max, produced during the described maneuvers, can
provide specific information about maximal diaphragm strength.
However, these tests require passage of balloon catheters and
the necessary coordination is difficult for naive subjects and
patients. There are limited normal data. It is difficult to control for muscle (fiber) length, and for velocity of shortening.
This test is recommended only as a research tool or in respiratory muscle function laboratories with specialized expertise.
Sniff Tests
Figure 4. Measurement of maximal static respiratory pressures. A

flanged mouthpiece with a nose clip is the preferred technique. A
small leak is introduced into the system, and a valve system allows a
normal breath to be followed by a maximum maneuver.
loon catheter systems passed into the esophagus (see TECHNIQUES FOR PRESSURE MEASUREMENTS) to measure Pes as a reflection of Ppl or into the stomach where Pga can be used to
reflect Pab. Esophageal pressure does not include lung elastic
recoil pressure but does include chest wall recoil pressure. The
main indication for balloon catheter measurements of maximum respiratory muscle pressures is to estimate the strength
of the separate muscle groups, notably the diaphragm (from
Pdi), or to measure strength when the patient is unable to
maintain a proper seal around the mouthpiece.
With the balloon catheters in place, various maneuvers can
be used to assess global inspiratory muscle or diaphragm
strength. These tests are usually performed at FRC. In the
Mueller (maximal inspiratory) maneuver the diaphragm and
inspiratory muscles are contracted with the aim of creating the
biggest negative thoracic pressure without regard to abdominal
pressure. However, this usually does not generate maximum
Pdi (25, 52). As an alternative, the subject may perform an expulsive maneuver, wherein the individual is requested to bear
down as for defecation and simultaneously superimposes a
Mueller maneuver. When given visual feedback, this complex
maneuver can be mastered by trained subjects to give the largest values of Pdi (up to 240 cm H2O or more) (53). It may reflect nearly maximal neural activation of the diaphragm, perhaps with fiber lengthening (52, 54). However, the technique is
difficult for naive subjects and in the clinical setting (55).
Twitch occlusion studies have confirmed that such maneuvers
can produce maximal neural activation of the diaphragm (56).
Scientific basis. A sniff is a short, sharp voluntary inspiratory

maneuver performed through one or both unoccluded nostrils. It involves contraction of the diaphragm and other inspiratory muscles. To be useful as a test of respiratory muscle
strength, sniffs need to be maximal, which is relatively easy for
most willing subjects, but may require some practice.
The sniff was described in 1927 as a radiological test of diaphragm paralysis because, in normal subjects, it was associated with crisp diaphragm descent during inspiration (57, 58).
Esau and coworkers (59) suggested that a short, sharp sniff
would approximate the diaphragm contraction elicited by a
brief stimulation of the phrenic nerves (59, 60). Miller and coworkers (61) showed that normal subjects generated greater
Pdi during maximal sniffs than during maximal static inspiratory efforts, perhaps because the maneuver achieves rapid,
fully coordinated recruitment of the inspiratory muscles (62).
The detailed respiratory mechanics of this dynamic maneuver
have been little studied, but numerous studies using the sniff
in normal subjects and patients have found it to be a robust
measure. The nose appears to act as a Starling resistor, so that
nasal flow is low and largely independent of driving pressure,
Pes (63). Pdi measured during a sniff (Pdi,sn,max) reflects diaphragm strength and Pes reflects the integrated pressure of
the inspiratory muscles on the lungs (Figure 6).
More recently it has been suggested that pressures measured in the mouth, nasopharynx, or one nostril give a clinically
useful approximation to esophageal pressure during sniffs (64,
65). Because these measurements do not require the passage
of esophageal or gastric balloons, they are easier for operator
and subject. However, pressure transmission may be impaired,
particularly when there is significant disease of the lungs (66).
Methodology. For measurement of maximal sniff pressures, patients are encouraged to make maximum efforts.
Sniffs can be achieved only when one or both nostrils are unoccluded, to allow the passage of air. An occluded sniff may
be called a gasp, and is more difficult for subjects to perform
reproducibly. Subjects should be instructed to sit or stand
comfortably, and to make sniffs using maximal effort starting
Figure 5. (A) A typical pressure tracing from a subject performing a maximum expiratory maneuver (PEmax). A peak
pressure is seen and the 1-second average is determined
by calculating the shaded area. (B) A pressure tracing from
a subject performing a maximum inspiratory maneuver
(PImax).
534
from relaxed end expiration. Detailed instruction on how to

perform the maneuver is not necessary, and may be counterproductive. However, subjects should be exhorted to make
maximal efforts, with a rest between sniffs. Most subjects
achieve a plateau of pressure values within 510 attempts.
Transdiaphragmatic pressure during sniff. Esophageal and
gastric balloons are passed by the usual technique. Transdiaphragmatic pressure during sniff (Pdi,sn) is reasonably reproducible within normal subjects, although there is wide variability between subjects (61) (Table 3). The values tend to be
as large, or larger, than Pdi during maximum static inspiratory
efforts (61).
Esophageal pressure during sniff. The methodology is as for
Pdi,sn but with the passage of an esophageal balloon alone.
Nasal sniff pressure. Pressure is measured by wedging a catheter in one nostril. Various techniques for wedging are available including foam, rubber bungs, and dental impression
molding. The subject sniffs through the contralateral unobstructed nostril. The pressure in the obstructed nostril reflects
the pressure in the nasopharynx, which is a reasonable indication of alveolar pressure. This in turn approximates esophageal pressure, particularly if the lungs are normal with a
mean Pnas/Pes ratio of 0.92 (64, 66). In COPD, nasal sniff
pressure (Pnas,sn) tends to underestimate esophageal pressure during sniff (Pes,sn) but can complement PImax in excluding weakness clinically (67).
Mouth and nasopharyngeal pressures can also be measured, and also reflect alveolar pressure, but are less easy for
the subject than nasal pressure and have no significant advantages (65) (Figure 7).
The sniff is a dynamic maneuver and so a pressure measurement system is required with a frequency response of
10 Hz. This can be achieved by a standard balloon catheter
system with a suitable pressure transducer. Use of cathetermounted transducers has also been described (18).
Advantages. The sniff is easily performed by most subjects
and patients and requires little practice. It is relatively reproducible and has a smaller range of normal values than mouth
pressures (61, 62). It is a useful voluntary test for evaluating
diaphragm strength in the clinical setting (55), giving equal or
greater pressures than maximal static efforts (61, 68).
It is possible to achieve greater transdiaphragmatic pressures by certain maneuvers, such as the modified Mueller maneuver (see previous passages), in highly trained and wellmotivated subjects. This may be important in physiological
studies but is not usually clinically relevant.
Sniff nasal pressure is technically simple.
Disadvantages. The pressures measured during a sniff may
be less than maximal static values because of shortening of the
inspiratory muscles (pressurevelocity relationship) (69). The
average volume change during a sniff is approximately 500 ml
with some gas rarefaction, which may be somewhat more than
the volume by which gas expands during a static maneuver
against a closed airway (63).
Sniffs are difficult or impossible if there is upper airway distortion, and particularly if the nose is completely obstructed.
It may be difficult to pass the balloons if there is severe bulbar
weakness, but this would not preclude measurement of Pnas.
Sniffs are voluntary maneuvers and, therefore, poorly motivated subjects may perform submaximal efforts. These can
often, but not always, be detected as variability tends to be
greater than for maximal maneuvers.
The sniff generates rapid pressure changes, so measurement requires a catheter system and transducer with a higher
frequency response (see previous passages) than for static maneuvers.
VOL 166
2002
Figure 6. Esophageal (Poes), gastric (Pg), and transdiaphragmatic (Pdi)

pressures measured during maximum voluntary sniffs in a normal subject and in a patient with severe diaphragm weakness. The normal subject reproducibly generates a Pdi of 120 cm H2O (11.8 kPa), whereas
the weak patient can generate only 15 cm H2O (1.5 kPa).
Applications. Maximal sniffs have been widely used and

validated as reproducible and reliable tests of diaphragm or
global inspiratory muscle function. This test can, therefore, be
used in research studies, although care must be taken with reproducibility, as in any voluntary maneuver.
The sniff maneuver is a useful part of the clinical evaluation of respiratory muscle strength and correlates well with response to therapy (Figure 6) (68). The normal values are
shown in Table 3 (61). There is a wide normal range, reflecting
the wide range of normal muscle strength in different individuals. In clinical practice Pdi,sn,max values greater than 100 cm
H2O in males and 80 cm H2O in females are unlikely to be associated with clinically significant diaphragm weakness (55).
Values of maximal sniff Pes or Pnas numerically greater than
70 cm H2O (males) or 60 cm H2O (females) are also unlikely to be associated with significant inspiratory muscle
weakness. However, these reflect the integrated pressure of
all the inspiratory muscles and it is possible that there could be
a degree of weakness of one or more of these muscles that
would not be detected at this level.
Cough Tests
Scientific basis. Measurement of pressure during cough is of

interest because the main expiratory muscles, the abdominal
muscles, are also those used in cough. Reduced cough pressure is of clinical importance because it may predispose to
chest infections. Also, in some patients technical difficulties
preclude measurement of mouth pressure during a static maximal expiratory maneuver (PEmax); in these patients measurement of maximal cough pressures is an alternative measurement technique.
A normal cough consists of four phases: inspiratory, compressive, expulsive, and relaxation (70). For cough to be a
measurement of expiratory muscle strength, two aspects re-
TABLE 3. TRANSDIAPHRAGMATIC PRESSURES DURING MAXIMAL

STATIC RESPIRATORY EFFORTS AND MAXIMAL SNIFFS
PI,di,max
(cm H2O)
Men
Women
All
Pdi,sn
(cm H2O)
Mean
SD
Range
Mean
SD
Range
37
27
64
108
65
90
30
31
37
52164
1640
16164
148
121
137
24
25
28
112204
82182
82204
Definition of abbreviations: PI,di,max maximum static transdiaphragmatic pressure;

Pdi,sn transdiaphragmatic pressure during sniff.
535
quire consideration: how standard is the maneuver and what

should be measured.
Expiratory muscle strength is influenced by lung volume
(71). It is normal to take a deep breath before a maximal
cough. Thus, although no instructions are given concerning
the magnitude of inspiration, the actual lung volume is probably relatively constant for a given individual during serial measurements. Compression requires a functional glottis; in some
disorders, for example bulbar type amyotrophic lateral sclerosis, this may not be present (72). A glottis that does not open
immediately causes an uncomfortable choking sensation and
makes measurement of mouth pressure (during either a PEmax
maneuver or a cough) difficult, but would not exclude obtaining
useful measurements from a gastric or esophageal balloon (73).
Theoretically, the peak cough pressure could be measured
at the mouth, esophagus, or stomach, but measurements at the
mouth have not been reported. Pdi is generated during cough
(74) and forced expiration (75) so that Pes,co is always less
than Pga,co (76). This Pdi can be substantial in a few subjects
(76); it may be due to active contraction of the diaphragm, or
passive stretching. In general, gastric pressure can be viewed
as a reasonable measure of abdominal muscle strength. The
surface EMG from abdominal muscle also varies with cough
intensity (77), but difficulties in standardizing it between measurements make this relatively impractical as a measurement
of strength.
Methodology. After passage and positioning of appropriate balloon catheters the subject is asked to cough as forcefully as possible. Visual feedback seems to be helpful, as with
the diaphragm (53). Peak pressures are measured between the
baseline at relaxed end-expiratory lung volume, and peak pressure (Figure 8).
Voluntary coughs are usually initiated from above FRC.
Lung volume may need to be controlled or measured, although this may not be necessary for clinical measurements.
It would, theoretically, be possible to measure the pressure
generation during an induced (e.g., with citric acid) cough in
patients unable to cough voluntarily, for example, those in intensive care units.
Cough pressures can be large, so it is necessary to have an
adequate volume of air in the balloon, to avoid compression of
the gas bubble in the balloon and its displacement into the
cathetertransducer system.
Normal ranges. No normal data exist from large studies.
Black and Hyatt (78) found PEmax and cough esophageal pressures comparable in a subgroup of their subjects. Other authors have found lower values in patients with COPD (79).
The mean maximal Pga,co in a small group of normal subjects
2075 years of age was 230 cm H2O for men and 166 cm H2O
for women, with lower limits of 160 cm H2O for men and 120 cm
H2O for women (76).
PRESSURES OBTAINED VIA PHRENIC

NERVE STIMULATION
Scientific Basis
The diaphragm is innervated exclusively by the phrenic nerve

and thus phrenic nerve stimulation (PNS) provides a specific
means to investigate the diaphragm independent of other inspiratory muscles. Indices that may not be specific for diaphragm contraction when measured during voluntary maneuvers do relate to the diaphragm when they are derived from
PNS. Examples include surface recordings of the electromyogram of the costal diaphragm (EMGdi), Pmo or Pes, and
phonomyogram (PMGdi). The other major feature of PNS is
that it eliminates the influence of the central nervous system.
Figure 7. Simultaneous pressure measurements from esophageal

(Pes), nasopharyngeal (Pnp), and mouth (Pmo) balloons in a normal
subject performing sniffs of increasing strength, to illustrate close correlation of the waveforms. Reprinted by permission from Reference 65.
PNS can give important information about the mechanical

function of the diaphragm, namely, about how the force of
contraction is transformed into pressure, and can be used to
confirm whether a contraction is maximal. PNS superimposed
on naturally occurring or voluntary contractions (twitch occlusion principle; see subsequent section) can provide an objective estimate of the maximal voluntary pressure that the diaphragm can produce.
Methodology
Since 1980, PNS has been investigated as a technique for elucidating mechanical aspects of diaphragm function. The 1989
National Heart, Lung, and Blood Institute (NHLBI) workshop on respiratory muscle fatigue identified it as one of the
most promising techniques in this field (80).
The stimulus to the nerve, which can be an externally applied
electric field or secondary currents surrounding a magnetic field
(see MAGNETIC STIMULATION), elicits synchronized activation of
motor units and subsequent muscle contraction. The effects of
phrenic nerve stimulation can be studied both electrophysiologically (see STIMULATION TESTS in Section 3 of this Statement) and
mechanically (this Section).
Four main PNS techniques have been used, mostly in
healthy volunteers, less often in patients. Two of them, needle
stimulation (81, 82) and implanted wire stimulation (83), are
invasive, with the risk of hematoma and phrenic nerve damage. Needle stimulation is not now recommended, and is not
further discussed. Implanted wire stimulation is probably safer,
and may be a convenient means to obtain repeated twitches
over long periods of time. The two others techniques, transcutaneous electrical PNS (ES) and magnetic stimulation (MS),
have been more extensively studied and have minimal side effects.
Subjects in Phrenic Nerve Stimulation
For laboratory and clinical studies subjects should be sitting in

a comfortable chair. Headrests may be helpful for ES.
Lung volume. A major condition for evaluating the pressure response to PNS is adequate relaxation of the respiratory
muscles at FRC, or within approximately 500 ml (84).
Posture. Most studies have been performed with seated
subjects, but twitch transdiaphragmatic pressure (Pdi,tw) seems
to be little altered by posture in contrast to sniff transdiaphragmatic pressure (Pdi,sn) and static Pdi (56, 65, 84, 85).
This feature could be useful in an intensive care unit (ICU)
setting.
Abdominal binding. Abdominal binding has little effect on
voluntary Pdi values (86), but markedly increases Pdi,tw (56,
65, 84, 87). The rationale for binding the abdomen during PNS
is to make the contraction closer to isometric than with a com-
536
Figure 8. Pressures during a maximal voluntary cough in a normal subject showing high positive gastric pressures generated in abdomen
(Pga; thick line) and esophagus (Pes) with low Pdi during the maneuver.
pliant abdominal wall; this may be appropriate for physiological studies (56, 88, 89). In a clinical setting it is difficult to standardize a binding technique and so the abdomen is usually
unbound.
Transcutaneous Electrical Phrenic Nerve Stimulation
Scientific basis. An externally applied electrical field induces

depolarization of phrenic nerve fibers, at mid-distance between the electrode and the cathode. If the stimulus is intense
enough, all fibers are activated synchronously, giving predictable and reproducible results (see STIMULATION TECHNIQUES in
Section 3 of this Statement).
Methodology. For bilateral ES, the operator should stand
behind the seated subject. The skin in the stimulated region is
degreased and mildly abraded to decrease its electrical impedance, allowing lower current intensities. Monopolar or bipolar
electrodes can be chosen (Figure 9). For monopolar electrodes, the anode is usually taped on the skin below the clavicle
medially, and the cathode is held in the hand. Monopolar electrodes probably make it easier to find the nerve because a
greater number of spots can be tested. However, because the
electrical field is less focused than with bipolar electrodes, it
may be more difficult to avoid costimulation of the adjacent
sternocleidomastoid muscle or the brachial plexus. Bipolar
electrodes are more specific but are also slightly more difficult
to use. Various models of bipolar electrode are commercially
available. They generally include felt tips 5 mm in diameter
with an interelectrode distance of approximately 2 cm. When
performing stimulation, the tips of the electrode should be
soaked in saline and the cathode should be proximal.
The phrenic nerve is usually located beneath the posterior
border of the sternocleidomastoid muscle, at the level of the
cricoid cartilage (Figure 9). It is easier to locate and to isolate
from the brachial plexus in subjects with long and slender
necks. A simple technique to locate the nerve is to set the
stimulator on repetitive stimulation mode, e.g., at a frequency
of 1 Hz, with a relatively low intensity, and to try various sites.
Identification of the correct site may be aided by careful observation of the abdomen; it will therefore be desirable to remove the shirt. Once the nerve is identified, the operator
marks the spot and the orientation of the electrode. Current
intensity is then increased, while monitoring the EMG to ob-
VOL 166
2002
tain supramaximal stimulation (see subsequent paragraphs),

which is generally achieved with 30- to 50-mA shocks. This
procedure is performed on each side separately, before applying ES bilaterally. It is then advisable to reconfirm that stimulation is supramaximal. It is also possible to judge supramaximality from the plateau of pressure response, but changes can
be more difficult to interpret unless the plateau procedure is
repeated to check that there is no loss of supramaximality.
Equipment. A constant-current stimulator capable of delivering square wave shocks of 0.1-millisecond duration and of
modulable intensity is used, which should include two synchronized outputs to allow bilateral stimulation. Two triggered EMG
amplifiers and a display should be available, so that the muscle
action potential (M wave) can be checked online by the operator.
Several manufacturers provide complete machines that offer a
wide panel of sophisticated stimulation and EMG acquisition options. Stimulators and amplifiers can also be bought separately.
Advantages. If skillfully performed, ES generates a pure
diaphragmatic contraction. The corresponding output is thus
representative of diaphragm properties alone. ES can be reproducible in skilled hands.
Disadvantages. The stimulus intensities required to achieve
supramaximal stimulation can be uncomfortable. From the technical point of view, there are several difficulties. First, maintaining optimal contact between the stimulating electrode and
the nerve can be difficult. It may be necessary to impose a significant degree of pressure on the soft tissues of the neck,
which can be painful for the subject and awkward for the operator, particularly in obese or old subjects, or those with hypertrophy of neck muscles. Skin-taped stimulating electrodes
have been used in healthy volunteers (90), but they probably
do not guarantee reproducible results in all settings. Neckand electrode-stabilizing devices have been proposed (10, 85)
that can be effective, but are cumbersome. Second, it is sometimes impossible to dissociate PNS from brachial plexus stimulation, particularly at high current intensities. This can be a
source of discomfort for the subject, and can theoretically
modify the characteristics of the rib cage, with which the contraction of the diaphragm interacts. Third, it can be impossible
to locate the nerve, or to do so easily enough to obtain reliable
supramaximal stimulation. Because of these difficulties, maintaining a constant symmetric maximal stimulus may need repetitive ES, which in itself can increase twitch pressure by
potentiation or the staircase phenomenon (91, 92) (see also
TWITCH POTENTIATION).
The technical expertise required for effective ES may, thus,
be a source of variability in research studies and limit its use in
the clinical field, particularly in demanding settings such as the
ICU or exercise.
Magnetic Stimulation
Scientific basis. The ability of magnetic fields to stimulate nervous structures has long been known (93). Magnetic stimulation creates intense and brief magnetic fields, which, unlike
electric currents, are only mildly attenuated by natural barriers such as skin and bone. They can therefore reach deep nervous structures, where stimulation is produced in situ by the
electrical fields induced by the rapidly changing magnetic
fields (Figure 10). The mechanisms of neural response to magnetic stimulation are different from those of the response to
electrical stimulation (9497), and therefore the results obtained with the two techniques may have different interpretations. Nevertheless, magnetic stimulation has the advantage of
being relatively painless and is thus easily applicable in the
clinical setting. Several review articles were discussed by
Chokroverty (98).
537
Figure 9. Technique for transcutaneous electric stimulation of the

phrenic nerve. (A) Use of a bipolar electrode to locate and stimulate
the phrenic nerve; (B) the monopolar technique. The phrenic nerve is
usually found underneath the posterior border of the sternocleidomastoid muscle, at the level of the cricoid cartilage. The operator stands or
sits beside (possible for unilateral stimulation) or behind (for bilateral
stimulation) the patient and uses the electrode to push the muscle forward. Firm pressure on the soft tissues of the neck avoids changes in
the relationship between the electrode and the nerve.
During the last 10 years, magnetic stimulation has been extensively used to stimulate the central nervous system in conscious humans (99). From the respiratory point of view, magnetic stimulation applied to the cervical spine (CMS) elicits a
bilateral diaphragm contraction (100). The coil is centered
over the spinous process of the seventh vertebral body (C7),
but this does not mean that the seventh roots are stimulated.
Depolarization of a nervous structure by magnetic stimulation
requires that the stimulating current and the nerve share a
common pathway: centering a circular 90-mm coil around C7
would, thus, generally stimulate the third to fifth cervical roots
(101103), depolarized in their intraforaminal segment (104,
105). Although it is generally believed that CMS provokes diaphragm contraction through the stimulation of cervical roots,
it has been suggested that the C7 CMS magnetic field may
reach the phrenic nerves anteriorly, through the neck, and
thus stimulate the phrenic nerve trunk at a point more distal
than with ES (106) (Figure 11).
Cervical magnetic stimulation also stimulates other elements of the cervical roots and nearby nerves, thus causing
some contraction of neck and upper rib cage muscles, as well
as diaphragm (100, 107109) (see COMPARISON BETWEEN TRANSCUTANEOUS ELECTRICAL PHRENIC NERVE STIMULATION AND CERVICAL MAGNETIC STIMULATION).
Methodology. The subject, comfortably seated in a chair, is
asked to bend the neck forward slightly. The coil is applied to
the back of the neck, its midline coinciding with the axis of the
vertebral column (Figure 11). Optimal results are generally
obtained with the coil centered around the spinous process of
the seventh cervical vertebra (C7), but slightly higher and
lower positions should be tried with monitoring of pressure or
EMG, although care may be required to obtain satisfactory
surface EMG signals. The optimal coil position may vary with
the size and neck morphology of the subject. Stimulation intensity is generally set to the maximal output of the stimulator
(see SUPRAMAXIMAL STIMULATION).
Several manufacturers provide magnetic stimulators suitable
for CMS. High-powered machines, capable of producing mag-
Figure 10. Schematic representation of the principle

of magnetic stimulation. The
electric current produced by
the charging circuit is stored
in capacitors. Switching on
the trigger circuit results in
a sudden pulse of electric
current in the coil of wire
held in contact with the patient. The intensity of this
pulse of current can be modulated (from 0 to 100% of
maximal intensity). A pulsed
magnetic field orthogonal to
the current flowing through
the coil is produced. This
magnetic field is able to penetrate body tissues such as skin or bone with little attenuation. It in
turn induces secondary electric currents of low intensity in deep structures. If such a structure, for example, the cerebral cortex or the trunk
of a peripheral nerve, lies at a tangent to the secondary electric currents, it is depolarized at a point that depends on the local geometry
and on the intensity of the current. The intensity of the magnetic field
induced within the tissues decreases with distance. The maximum magnetic field intensity produced by modern stimulators is around 2.5 Tesla.
netic fields of 22.5 T with a medium-size circular coil, should

be used. Doughnut-shaped coils 90 mm in diameter are particularly suitable for generating bilateral diaphragmatic contractions for the measurement of twitch pressure (100, 107110).
Advantages. Cervical magnetic stimulation provides easy
bilateral PNS. It is not painful: The subject simply perceives a
contraction of neck muscles that provoke an extension movement, and a hiccup-like sensation. Any operator reasonably familiar with medical or physiological tests can obtain reliable
results after a brief period of training. The number of stimulations applied during a given CMS session is often lower than
with ES, which reduces the risk of potentiation and the staircase phenomenon (see previous passages, and TWITCH POTENTIATION, subsequent section). Location of the nerves is technically easier with CMS. The risk of falsely low results due to
difficulty in locating the nerves and other technical problems is
lower with CMS than with ES. In addition, because of its sites
of action (cervical roots of the phrenic nerve [101, 103] or intramediastinal segment of the phrenic nerve [106]), CMS can activate diaphragm fibers innervated by an accessory or ectopic
phrenic nerve (111) that would not be accessed by ES (112).
Disadvantages. Cervical magnetic stimulation lacks the specificity of ES for the diaphragm, because of coactivation of
muscles innervated by cervical roots or the brachial plexus. Interpretation of Pdi,tw by CMS is, therefore, not exactly the
same as that of Pdi,tw by ES (see subsequent section). Confirming supramaximal stimulation can sometimes be difficult
(see subsequent section). Obtaining a reliable EMGdi signal is
difficult with CMS, but technical solutions are available (use
of shielded EMG cables, transient muting of the EMG amplifiers, etc.) (113) and modern EMG recorders are now designed to support magnetic stimulation. A reasonable distance
should be maintained between the stimulating coil and credit
cards, computer disks, and the like.
Other Magnetic Stimulation Techniques
Focal magnetic phrenic nerve stimulation. Small figure of eightshaped magnetic coils can be used for focal stimulation (focal
MS) of the phrenic nerve in the neck unilaterally or bilaterally,
at the same point as stimulation by ES (114, 115). Bilateral focal
MS is easily applied by an operator standing in front of the sub-
538
ject and gives values for Pdi,tw and the Pga,tw/Pes,tw ratio that
are close to ES (115), thus avoiding any problems associated
with stimulating upper trunk muscles (see subsequent section).
This technique could make bilateral PNS easier in supine patients, as in the ICU, because for ES the operator has to stand
behind the patient and for CMS the subjects neck has to lie over
the coil, which may be uncomfortable and impede optimal positioning. Unilateral focal MS may allow assessment of the mechanical properties of one hemidiaphragm alone (87, 114, 116).
Anterior magnetic stimulation. The possibility of evoking a
bilateral diaphragm EMG response through anterior magnetic
stimulation (antMS) with a 90-mm circular coil similar to that used
for CMS, placed flat over the upper part of the sternum, has been
described (106). Anterior magnetic stimulation is potentially a
simple technique also applicable to supine subjects in difficult
settings, but pressure responses remain to be evaluated.
VOL 166
2002
phragm properties, but also on the load the diaphragm acts

against, as for any muscle. This load depends on the mechanical characteristics of the rib cage and abdominal wall (see
TECHNIQUES FOR PRESSURE MEASUREMENTS).
Measurement of the sounds created by muscle contraction
can now be quantified by phonomyography. At present this is
a research tool, but it has considerable promise as it is technically easy and noninvasive (117).
Transdiaphragmatic pressure. In response to PNS, Pdi rapidly rises to a peak, and then decreases exponentially to its
baseline value (Figure 12) to give a characteristic Pdi,tw. The
Diaphragmatic Response to Phrenic Nerve Stimulation
Diaphragmatic response to PNS would ideally be measured as

work or power. However, length changes and velocity of
shortening are invariably ignored and the focus has been on
assessing diaphragm force development, either by measurement of pressure, or sound (phonomyography).
Pressure responses to PNS (twitches) are widely used to
study diaphragm contraction. They depend not only on dia-
Figure 11. Technique for

cervical magnetic stimulation (CMS). Top: The usual
technique for CMS. The
subject being seated, a circular coil of approximately
90 mm in diameter attached to a magnetic stimulator is centered on the
spinous process of the seventh cervical vertebra (C7).
The subject is usually asked
to bend the neck forward
to facilitate contact between
the coil and the posterior
surface of the neck. In this
example, esophageal and
gastric pressures (Pes and
Pga, respectively) are measured via conventional balloon-catheter systems (see
TECHNIQUES FOR PRESSURE MEASUREMENTS). Subtracting Pes
from Pga gives transdiaphragmatic pressure (Pdi).
Surface electrodes are shown
for the assessment of the
electromyographic response
of the diaphragm to stimulation (EMGdi). Bottom:
Sites of nervous stimulation
possibly responsible for the
diaphragmatic contraction
induced by CMS using a
conventional 90-mm doughnut-shaped coil. The gray
area is a schematic representation of the magnetic
field. This can, theoretically,
depolarize cervical roots, and also the phrenic nerve itself anteriorly,
behind or below the clavicle. Phrenic nerve stimulation is theoretically
possible with a circular coil placed anteriorly (lying flat on the upper
part of the sternum: anterior magnetic stimulation) or posteriorly but
lower (coil held vertical on the upper dorsal vertebral column).
Figure 12. Typical pressure tracings in response to phrenic nerve stimulation. Traces A to C show, respectively, the esophageal pressure
(Pes), gastric pressure (Pga), and transdiaphragmatic pressure (Pdi) responses to bilateral, supramaximal, electrical stimulation of the phrenic
nerve in the neck, in a patient with COPD (hence the relatively low
amplitudes). On the Pdi trace (C) are indicated some indexes often
used to describe quantitatively the twitch response: (1) Pdi,tw is the
amplitude of the response from baseline to peak; it is the result of the
interaction of diaphragm contraction with the rib cage and the abdomen; although not a direct measure of diaphragm intrinsic contractile
properties, it is related to diaphragm strength and can be used to
characterize it if all other intervening factors are otherwise kept identical (e.g., lung volume, thoracic geometry, rib cage and abdomen
compliance, diaphragm state of activation, etc.); (2) ttp and 1/2rt are
the time-to-peak and half-relaxation time, respectively; these indexes
are used to characterize the dynamics of diaphragm contraction, and
are influenced, for example, by muscle shortening or fatigue; and (3)
is the time constant of an exponential function fitted to the after-peak
decline in Pdi (often referred to as the relaxation time constant); it is
influenced by diaphragm intrinsic properties and, for example, is prolonged by fatigue. The trace in D illustrates the similarity in shape,
time dynamics, and amplitude in the Pes twitch (thin line) and the
mouth pressure (Pmo) twitch (thick line) to phrenic nerve stimulation,
in a normal healthy volunteer.
539
time-to-peak and its first time derivative (rate of rise, dp/dt)

depend on several factors, including previous muscle shortening (e.g., increase in lung volume). The amplitude of the
twitch reflects the transformation of diaphragm force into
pressure and depends on diaphragm strength and contractile
properties as well as rib cage and abdominal wall compliance.
The dynamics of relaxation of the twitch can be described by
the time necessary to reach a Pdi value of 50% of the peak
(half-relaxation time) or by the time constant () of an exponential fitted to the pressuretime relationship (87, 118) (see
RELAXATION RATE in Section 5 of this Statement).
Mouth pressure. Provided that the diaphragm contracts in
isolation and that the corresponding change in alveolar pressure is adequately transmitted to the airways opening, Pmo
can in theory reflect diaphragm contraction. In response to PNS,
a twitch-shaped negative Pmo swing is seen (Pmo,tw) (Figure
12). In healthy subjects, Pmo,tw closely matches Pes,tw at different lung volumes and correlates with Pdi,tw (119). Conversely, in patients with COPD, Pmo,tw at relaxed FRC is
damped and time lagged with respect to Pes,tw, due to an increased airway time constant (120). Both in normal subjects
(110) and in patients with diaphragm weakness (121) there is
adequate matching of Pmo,tw and Pes,tw with CMS, when
precautions are taken to prevent glottic closure. In normal
subjects, Pmo,tw measured during CMS is generally more negative than 11 cm H2O (110), depending on lung gas volume (VL).
Although there are not yet enough data to propose a precise technique to measure Pmo,tw, the following recommendations seem reasonable. Stimulation should be attempted at
relaxed FRC when respiratory system recoil pressure is ordinarily zero. When PEEPi is present, the twitch amplitude
should be measured starting at the PEEPi level. If a supramaximal stimulation is obtained, the data can be retained, if
time to peak tension is normal. If it is prolonged with low amplitude and prolonged relaxation, then abnormal pressure
transmission (probably due to glottic closure) should be first
suspected, and stimulation should be repeated during a mild
expiratory effort against an occluded mouthpiece.
Advantages. The principal advantage of Pmo,tw is its simplicity and ease of use. Thus, portable Pmo or Psn devices (28)
could probably be adapted for combination with CMS as a
simple screening test.
Disadvantages. The main disadvantage of Pmo,tw is in ensuring the adequacy of pressure transmission from the alveoli
to the mouth, particularly in airway obstruction. It seems that
glottic closure, which may prevent change in Pmo, is particularly frequent with CMS, although it can occur with ES as
well. A technique such as a mild inspiratory effort at FRC
(120) or an expiratory effort (110) not only makes the procedure more complicated, but it also changes the meaning of the
observed results. Indeed, central nervous system activation
and lung volume influence the pressure response to PNS (see
CONFOUNDING FACTORS). With CMS, Pmo,tw can be the product of diaphragm contraction but also of neck muscle contraction. The importance of this confounding factor remains to be
studied in detail (see CONFOUNDING FACTORS).
Esophageal Pressure
The measurement of esophageal pressure alone in response to

PNS provides an intermediate approach between the measurement of Pdi and Pmo. Measuring Pes,tw is less simple
than measuring Pmo,tw, but the problem of incomplete pressure equilibration due to glottic closure or airways time constant does not exist. Normal values for Pes,tw are by definition
similar to values for Pmo,tw.
Comparison between Transcutaneous Electrical Phrenic Nerve

Stimulation and Cervical Magnetic Stimulation
Practical aspects. The practical aspects of ES and CMS are

described in previous sections. In summary:
1. ES is the original method for generating an isolated contraction of the diaphragm, but for the operator is difficult
to master.
2. CMS is easier and faster to apply, with a lower risk of false
results due to technical problems. It is better tolerated by
the subject. There is some cocontraction of the upper rib
cage and neck muscles, stiffening the rib cage, so that Pdi
may be greater than with ES.
Physiological aspects. The two techniques are comparable
with respect to reproducibility (similar within-occasion and
between-occasion coefficients of variation) (100, 107) and the
time characteristics of the Pdi twitches are very close. When
supramaximal bilateral ES and CMS are compared in the
same subjects (107109) Pdi,tw values measured during CMS
appear consistently higher by approximately 2025% than
Pdi,tw measured during ES, the difference being accounted
for by more negative Pes values.
Conclusion and perspectives. Cervical magnetic stimulation
and ES do not provide the same physiological information, so
that normal values for PNS-related pressures (see APPLICATIONS AND PERSPECTIVES) depend on the technique used. This
also means that the results of CMS and ES may be affected
differently in diseases affecting both diaphragm and rib cage
muscles. The possibility of focused phrenic nerve magnetic
stimulation in the neck with small figure-of-eight coils bilaterally (114, 115) will probably reconcile the physiological need
for pure diaphragm contraction and the clinical need for simplicity. CMS and other techniques of magnetic stimulation of
the phrenic nerve, being easier to use than ES, should then
overcome the limitations of ES for large series, exercise, ICU,
or intraoperative studies.
Confounding Factors
Supramaximal stimulation. To provide valid information about

the maximal strength of the whole diaphragm, PNS should be
bilateral and supramaximal.
Although PNS should be supramaximal if Ptw is to accurately reflect maximal diaphragm mechanical output, this is
not necessary when PNS is used to study phrenic conduction
time (see Section 3 of this Statement). Failure to achieve supramaximality leads to underestimation of diaphragm mechanical output, overestimation of central drive when using
twitch occlusion (see PARTICULAR TECHNIQUES), and variability. The gold standard to achieve a reasonable degree of certainty about the supramaximality of PNS is to establish a recruitment curve using the EMGdi, and stimulation is maximal
when there is no further increase in EMGdi in response to an
increase in intensity. Increasing stimulus intensity by a further
1020% provides a reasonable safety margin that compensates for slight changes in the quality of the stimulus. With
CMS, it can be difficult to reach a clear plateau in amplitude
of the action potential with intensity, but the peak-to-peak
amplitudes of the action potentials produced by CMS at the
maximal intensity of stimulation with a 2.5-T magnet are not
different from the peak-to-peak amplitudes produced by supramaximal bilateral ES (109). The increasing power of stimulators may help reduce this problem in the future.
Lung volume. Lung volume has a major influence on the
ability of the diaphragm to produce pressure, during voluntary
static or dynamic maneuvers (13, 36, 122128) and in response
540
to PNS (83, 84, 117, 119, 129134). This is a result of the inverse relationships of length with force in skeletal muscles,
and of lung volume with diaphragm length (36, 69, 125).
Pdi,tw decreases as lung volume increases, with a prominent
reduction in Pes,tw that is close to zero at TLC (83, 129131,
134136) (Figure 13). Isovolume changes in rib cage and abdominal configuration also influence Pdi,tw (83, 84, 125).
A long-standing increase in lung volume tends to be compensated for by adaptive mechanisms at the level of the sarcomere, known as length adaptation in animals (137141) and
probably in humans (131, 142). Thus, the observations made
during acute changes in lung volume may not be as relevant to
chronic hyperinflation.
How sensitive to changes in lung volume is the pressure response to PNS? Between FRC and TLC, Pdi,tw and Pmo,tw decrease by approximately 3%/100 ml (83, 84, 119, 130, 132), and
between RV and FRC by approximately 5%/100 ml (83, 128).
These changes appear to be reduced if care is taken to avoid potentiation (129) and may be less in the elderly (131, 134).
Lung volume, and if possible rib cage/abdominal configuration, should be carefully controlled when assessing PNS pressures in research settings. When PNS is repeated in patients
with labile lung volumes, FRC should be measured on the day
of the study. Assessing lung volume may be less crucial for
clinical assessment, recognizing that a change in Pdi,tw or
Pmo,tw can reflect changes in diaphragm properties, or lung
volume, or both.
Twitch potentiation. A transient increase in the contractility of a skeletal muscle follows its contraction. This phenomenon is called potentiation (143, 144).
The possibility of twitch potentiation should be taken into
account when interpreting studies involving PNS. A period of
quiet breathing, e.g., 15 min, should be allowed before recording diaphragm twitches, particularly if maximal maneuvers or
sniffs are performed beforehand (145147).
Hypertrophy of neck muscles. When there is a bilateral
paralysis, Pes,tw and Pdi,tw during ES are zero. With CMS,
coactivated neck muscles can, theoretically, produce some degree of Pes,tw during CMS. This effect may be small in most
subjects (148), but could be larger in patients with hypertrophied inspiratory neck muscles (149).
VOL 166
2002
mal it completely suppresses the twitch (150153) (Figure 14).

It is usual to compare the amplitude of the interpolated twitch
with that of the fully potentiated twitch performed on the relaxed muscle.
Normal subjects can voluntarily produce maximal diaphragmatic contraction during inspiratory and expiratory efforts as
judged by twitch occlusion (56, 154, 155), as can patients with
COPD (131). However, interpolation can be complicated by
the effects of series compliance on twitch height (156).
Main results and applications. Superimposed Pdi,tw decreases
linearly with the degree of underlying voluntary Pdi (Pdi,vol) (Figure 14) according to:
Pdi,tw = a b Pdi,vol
(1)
The y intercept (a) of Equation 1 closely matches the value of

Pdi,tw obtained at relaxed FRC, and its x intercept closely
matches Pdi,max. The a/b ratio corresponds to the diaphragm
twitch-to-tetanus ratio, which is 0.200.25 in most studies,
close to the twitch-to-tetanus ratio of skeletal muscles in mammals (157), giving credibility to the technique.
From the twitch occlusion technique, it has been possible to:
1. Assess the degree of central activation associated with voluntary diaphragm contraction (88, 158).
2. Discriminate between the central and peripheral components of diaphragm fatigue or weakness (88, 159, 160) (see
TYPES OF FATIGUE in Section 5 of this Statement).
Particular Techniques
Twitch occlusion. Scientific basis. The degree of activation of

a skeletal muscle during voluntary efforts can be assessed by
use of the twitch occlusion technique. Twitches produced by
electrical stimulation of a parent nerve are superimposed on a
voluntary isometric contraction of the muscle (150). The amplitude of the twitch decreases linearly as the strength of the
underlying contraction increases. When the activation is maxi-
Figure 13. The effects of lung

volume on twitch pressure.
Shown is the marked decrease
in the amplitude of transdiaphragmatic pressure (Pdi,tw)
during an acute inflation from
functional residual capacity
to total lung capacity, both
in normal volunteers (solid
circles) and in patients with
preexisting hyperinflation impairing baseline diaphragm
efficacy (open circles). This reduction in Pdi with volume is mainly due to a reduction in the esophageal pressure (Pes) component of the twitch, which becomes very
close to zero at TLC in normal subjects.
Figure 14. Twitch occlusion technique. Top: Procedure that can be

used to study the effects of a voluntary diaphragmatic contraction
(Pdi,vol) on the response to phrenic nerve stimulation (Pdi,tw). Pdi is
displayed on an oscilloscope visible to the subject and the operator.
The subject is asked to perform stepwise graded contractions, guided
by the Pdi trace, and to sustain several levels of Pdi for a few seconds.
The operator superimposes one or more stimulations during each sustained contraction. Bottom: two examples of twitch occlusion procedures. In both cases, Pdi,tw decreases linearly with Pdi,vol, according
to Pdi,tw a b Pdi,vol. The left panel illustrates a subject capable of
maximally activating the diaphragm: the subject can produce a voluntary contraction that is associated with complete absence of Pdi,tw.
The right panel illustrates a subject not capable of maximal contraction: Pdi,max can be estimated by extrapolating the Pdi,twPdi,vol relationship to the x axis, and is equal to the ratio of the intercept of this
relationship to its slope (Pdi,max a/b, arrow).
541
3. Detect central inhibition of the drive to breathe (131, 161,

162).
4. Estimate Pdi,max from submaximal efforts (56).
Methodology. The technique by which bilateral supramaximal PNS is to be performed should be that established for ES
(56, 88, 120, 131, 154, 155, 160, 161, 163167) or CMS (108). A
visual feedback of Pdi is provided to the subject and the operator on an oscilloscope or computer screen. A standard procedure is as follows. First, Pdi,max is determined. Then, the subject is asked to produce, in a stepwise manner, fractions of
Pdi,max previously marked on the screen of the oscilloscope.
While a given step is briefly sustained, one or a few stimuli are
delivered so that data for Equation 1 can be built up. To assess
the degree of central activation a single stimulus over any inspiratory effort followed by a stimulus in relaxed condition
may be sufficient (88).
Advantages. Twitch occlusion allows an investigator to separate central from peripheral weakness or fatigue of the diaphragm (80).
Disadvantages. The twitch occlusion technique is restricted
by the difficulties inherent in ES and Pdi measurement. The
consistency of the stimulus delivered to the nerves becomes increasingly difficult to maintain as neck muscles are recruited
during intense inspiratory maneuvers, although CMS (109) or
Pmo (120) could help. When graded voluntary Pdi maneuvers
are performed, neck muscles often stay relaxed during lowintensity diaphragm contractions, but are coactivated during
high-intensity contractions (70% of Pdi,max and above) (56),
which may decrease rib cage distortability. During twitch occlusion from relaxed conditions to maximal effort, diaphragm
contraction will interact with a distortable rib cage at low
Pdi,vol values, and with a much stiffer rib cage at high Pdi,vol
values, complicating the results. This problem may disappear
with CMS, because it tends to stabilize the upper rib cage at all
levels of Pdi,vol (109). The impact of twitch potentiation on
the twitch occlusion procedure is not known. Theoretically,
potentiation of Pdi,tw would lead to an underestimate of the
central component.
Forcefrequency curves. When stimulation frequency increases, the mechanical output of the stimulated muscle increases
up to a plateau (tetanic contraction). Forcefrequency curves
contain much information about skeletal muscle contractility,
and reflect the type and severity of muscle fatigue.
Human diaphragm forcefrequency (or, rather, pressure
frequency) characteristics in vivo have been described with
unilateral PNS (81, 168, 169) and bilateral PNS (116, 170172),
in a limited number of highly motivated subjects. Despite their
interest, data remain scarce, because the studies are difficult.
First, sequences of supramaximal stimuli are painful and can
barely be tolerated, even unilaterally. Second, maintaining adequate sustained supramaximal PNS is demanding, especially
as uncomfortable spread of the stimulus to the brachial plexus
becomes almost unavoidable with repeated stimuli. Therefore,
the use of pressurefrequency curves appears limited and cannot reasonably be proposed for clinical purposes or even for
patient-based research. An alternative is to generate a surrogate forcefrequency curve using paired phrenic nerve stimuli
(173). This approach is acceptable to naive elderly subjects
(174) and patients (175).
Normal Values
Values for Pdi,tw or Pmo,tw in normal subjects can be found

in more than 40 published studies, irrespective of the technique used (10, 56, 65, 8385, 8890, 100, 107, 110, 115, 116,
118120, 129, 131, 132, 134, 145, 146, 155, 160, 161, 165167,
170173, 175190) (Table 4). Normal young (2035 years)

male individuals provide the vast majority of the data.
Bearing in mind that the technique used and various methodological factors can exert a major influence on the results, it
seems possible to propose the following, for bilateral PNS.
Amplitude of bilateral twitch transdiaphragmatic pressure.
Assuming a correct technique, at FRC, Pdi,tw with ES should
be above 15 cm H2O and Pdi,tw with CMS should be above 20
cm H2O.
A Pdi,tw below 15 cm H2O, whatever the PNS technique,
should raise high suspicion of diaphragm dysfunction (110,
114, 131, 179, 187, 190196).
Influence of age, sex, and other characteristics. As with voluntary pressures (26, 128, 197), normal Pdi,tw tends to be
lower in women than in men and tends to decrease with age
(120, 131, 182, 190). The influence of other factors such as
height, race, fitness, etc., is unknown, although data suggest
that baseline Pdi,tw in highly fit subjects is not different from
normal subjects (185), but with a better resistance to maximal
exercise-induced fatigue.
Other characteristics of the twitch transdiaphragmatic pressure response to phrenic nerve stimulation. The ratio of Pes,tw
to Pdi,tw in normal individuals is usually between 0.35 and
0.55.
Amplitude of twitch transdiaphragmatic pressure. Available data
are still scarce, but a value of Pmo,tw during CMS less negative than 11 cm H2O should probably prompt more detailed
investigation (110). Pmo,tw during ES tends to be lower, and a
value more negative than 8 cm H2O with this technique is
probably normal (119).
Applications and Perspectives
The use of PNS-derived pressures to study the mechanical action of the diaphragm assumes that stimulation is bilateral and
supramaximal, and that care is taken to control for lung volume and twitch potentiation.
Research applications. Bilateral ES is an important tool in
human diaphragm research, used to study the properties of the
diaphragm and the mechanisms of its fatigue independently of
volitional influences. The twitch occlusion technique, as the
only means to separate the peripheral and central components
of diaphragm dysfunction, is also an important, if complex, research tool. The use of paired stimuli could facilitate research
on the frequency characteristics of diaphragm fatigue (174).
Cervical magnetic stimulation provides slightly different information from ES, mainly because of rib cage muscle coactivation. However, its simplicity for the subject and the operator makes it a valuable tool for clinical research, especially in
difficult settings, such as the ICU, or when repeated studies
are needed, such as the evaluation of therapeutic interventions.
Clinical applications. Contraindications. There are virtually no contraindications to ES and MS, but their use requires
some precautions. MS should not be performed in patients with a
cardiac pacemaker. Patients with orthopedic implants, even
metallic ones, can be studied with MS (e.g., candidates for
phrenic pacing after high cervical cord injury) (198) after sufficient time for consolidation.
Clinical use. Pressure responses to PNS play an important
part in the clinical evaluation of inspiratory muscle function.
Diaphragmatic weakness can be established and quantified with
PNS, particularly when voluntary maneuvers are equivocal. The
use of MS with Pmo may be an even simpler nonvolitional
measure.
The preoperative assessment of patients for possible phrenic
pacing after high cervical cord lesions constitutes a particular
application. PNS provides information on phrenic nerve con-
542
VOL 166
2002
TABLE 4. TWITCH TRANSDIAPHRAGMATIC PRESSURES MEASURED IN A VARIETY OF STUDIES

First Author
(Ref.)
Pdi,tw
(cm H2O, mean)
Note
Bilateral Supramaximal ES: Healthy Young Volunteers

Bellemare (56)
Aubier (118)
Hubmayr (83)
Mier (179)
34.6
34.4
31.4
833
Mier (84)
Gandevia (155)
Mier (82)
24.4
28.0
19.429.3
Mador (90)
Wragg (107)
Eastwood (85)
28.9
29.7
2128
Laghi (108)
Similowski (109)
32.3
23.3
Similowski (131)
19.831.7
Needle stimulation
Implanted wire stimulation
15 control subjects referred for suspected diaphragm
weakness; significant overlap between groups; Pdi,tw was
discriminant only for diagnosis of severe diaphragm weakness
Supine position
3 subjects; comparison between transcutaneous electrical
stimulation and needle stimulation
Comparison CMSES
3 subjects; validation of a cervical apparatus aiming at
standardizing PNS for repeated studies: data reproducible
among 25 repeated studies
Comparison CMSES
Comparison CMSES
Bilateral Supramaximal ES: Older Healthy Volunteers
Mancini (182)
6 subjects aged 5072 yr, serving as control subjects in a study of

diaphragm properties in hyperinflated patients with COPD
6 subjects aged 50 8 yr, serving as control subjects in a study of
diaphragm function in chronic heart failure
18.8
Cervical Magnetic Stimulation: Healthy Young Volunteers

Similowski (100)
33.4
Wragg (107)
Wragg (146)
36.5
36.5
Laghi (184)
Hamnegard (129, 187)
Similowski (109)
Laghi (108)
First-generation stimulator; supramaximality of PNS not

always certain
Comparison CMES
Study of diaphragm twitch potentiation; values in the table are
the unpotentiated ones
12 subjects
38.9
1734
27.5
37.7
Comparison CMES
Comparison CMES
Cervical Magnetic Stimulation: Older Healthy Volunteers

Polkey (190)
25.4
Patients with COPD
Similowski (131)
Wanke (183)
Polkey (134, 190)
1026
15.2
18.2
Bilateral ES
Bilateral ES
CMS
Definition of abbreviations: CMS cervical magnetic stimulation; COPD chronic obstructive pulmonary disease; ES electrical stimulation; Pdi,tw twitch transdiaphragmatic pressure; PNS phrenic nerve stimulation.
Values represent means or ranges.
duction time, which is important in making the decision to undertake pacing (198). It also provides an estimate of the degree of diaphragm disuse atrophy, which is an important
determinant of the reconditioning strategy (199).
Given that methodological precautions are rigorously respected, PNS is one of the more reproducible respiratory muscle tests, and this makes it suitable for follow-up studies (see
PHRENIC NERVE STIMULATION in Section 3 of this Statement).
ABDOMINAL MUSCLE STIMULATION

The abdominal muscles are major contributors to respiration,
both through their expiratory action on the rib cage and their
mechanical linkage with the diaphragm. Their function can be
explored by voluntary maneuvers (see VOLITIONAL TESTS OF
RESPIRATORY MUSCLE STRENGTH), with the usual limitations of
subjects understanding and cooperation. A nonvolitional test
giving information on abdominal muscle mechanical output
would, therefore, be of interest.
Contraction of the abdominal muscles can be provoked by
their direct electrical stimulation. This technique has been
used in humans to study the action of various abdominal muscles on the rib cage (200), and in a study of diaphragm maximal voluntary activation (155). From a therapeutic point of
view, its use has been considered to enhance cough in patients
with cervical cord injury (201). However, direct electrical stimulation is painful and supramaximality is difficult to achieve. It
is also difficult to activate all muscles groups at once.
Contraction of the abdominal muscles can also be provoked by stimulation of their parent nerves and roots (202).
Theoretically, this allows supramaximal stimulation (and therefore reproducibility) and, if the stimulus is widespread enough,
simultaneous activation of all abdominal muscles. Magnetic
stimulation over the vertebral column at the level of the eighth
to tenth thoracic vertebra could provide an easy-to-use nonvolitional assessment of abdominal muscle strength and fatigue
(202, 203).
CONCLUSION
The purpose of this Section is to describe the tests used to assess respiratory muscle strength. To test strength, pressures
543
can be measured either during voluntary maneuvers or during

involuntary contractions, particularly in response to phrenic
nerve stimulation.
A. Volitional Tests of Respiratory Muscle Strength: Volitional tests are often simple for patients to perform, but it
can be difficult to be certain that a maximum effort has
been made. This can lead to difficulty in the interpretation
of low results.
1. Maximum Static Inspiratory and Expiratory Pressure
i. PImax and PEmax are commonly used, clinically useful
measurements. Some individuals have difficulty with
the technique and the interpretation of low results can
be problematic.
ii. Maximum static transdiaphragmatic pressure (PI,di,max)
provides specific information on diaphragm strength,
but can be a difficult maneuver in naive subjects and
patients. PI,di,max has a wide normal range and has
limited usefulness in clinical practice.
2. Maximum Sniff Pressures: Maximum sniff efforts can be
achieved by patients with little practice; sniff pressures
are reproducible and have a narrower normal range than
static mouth pressures or PI,di,max. Sniff esophageal
pressure assesses global inspiratory muscle strength and
sniff Pdi is a clinically useful measure of diaphragm
strength. Sniff nasal pressure provides a useful noninvasive measure of inspiratory muscle strength and has been
validated in patients with neuromuscular disease.
3. Maximum Cough Pressure: Cough gastric pressure is measured as an index of abdominal muscle strength. Pga,co is
a useful test to supplement PEmax, particularly in patients
unable to perform the PEmax maneuver reliably. To date,
few data are available for normal values of Pga,co.
B. Nonvolitional Tests of Respiratory Muscle Strength
1. Phrenic Nerve Stimulation: Phrenic nerve stimulation is
specific for the diaphragm and is not influenced by the
central nervous system.
i. Electrical phrenic nerve stimulation (ES) can achieve
selective supramaximal stimulation of the diaphragm,
but requires considerable skill, is sometimes uncomfortable for patients, and is difficult to achieve in
some clinical settings (e.g., the ICU).
ii. Magnetic phrenic nerve stimulation (MS) is technically easier for the operator and less uncomfortable
for the patients. Cervical magnetic stimulation (CMS)
elicits a bilateral diaphragm contraction. CMS is less
specific than ES, and coactivates muscles innervated
by the brachial plexus. Achieving and confirming supramaximal nerve stimulation can be difficult, and recording the diaphragm EMG can pose problems. Unilateral anteriorlateral MS is more specific than CMS,
and results are similar to ES. Unilateral MS allows the
investigation of hemidiaphragm function, and bilateral
anteriorlateral MS reliably achieves supramaximal
stimulation.
iii. Twitch transdiaphragmatic pressure (Pdi,tw) provides
an index of diaphragm, or hemidiaphragm, strength.
Normal values are available and Pdi,tw is a useful clinical measurement. Twitch mouth pressure (Pmo,tw)
can provide a noninvasive measure of diaphragm
strength, but inadequate transmission of alveolar pressure to the mouth, in patients with airway obstruction
or when there is glottic closure, is a substantial practical problem limiting clinical application.
iv. Pdi,tw is a research technique useful to assess the degree of activation of the diaphragm during voluntary
efforts (the technique of twitch occlusion) and Pdi,max

can be estimated from submaximal efforts.
2. Abdominal Muscle Stimulation: Magnetic stimulation
over the eighth to tenth thoracic vertebrae posteriorally,
and the recording of twitch gastric pressure, provide a
clinically applicable nonvolitional test of abdominal muscle strength. Data on normal values for twitch gastric pressure are limited.
References
1. Loring SH, Yoshino K, Kimball WR, Barnas GM. Gravitational and
shear-associated pressure gradients in the abdomen. J Appl Physiol
1994;77:13751382.
2. Loring SH, Kurachek SC, Wohl ME. Diaphragmatic excursion after
pleural sclerosis. Chest 1989;95:374378.
3. Chihara K, Kenyon CM, Macklem PT. Human rib cage distortability. J
Appl Physiol 1996;81:437447.
4. Maton B, Petitjean M, Cnockaert JC. Phonomyogram and electromyogram relationships with isometric force reinvestigated in man. Eur J
Appl Physiol 1990;60:194201.
5. Petitjean M, Bellemare F. Phonomyogram of the diaphragm during
unilateral and bilateral phrenic nerve stimulation and changes with
fatigue. Muscle Nerve 1994;17:12011209.
6. Milic-Emili J. Techniques in respiratory physiology. II. Measurements
of pressures in respiratory physiology. In: Otis AB, editor. Techniques in the life sciences: physiology, Vol. P4/II. New York: Elsevier; 1984. p. 122, 412.
7. Jackson AC, Vinegar A. A technique for measuring frequency response
of pressure, volume, and flow transducers. J Appl Physiol 1979;47:
462467.
8. Milic-Emili J, Mead JJ, Turner JM, Glauser EM. Improved technique
for estimating pleural pressure from esophageal balloons. J Appl
Physiol 1964;19:207211.
9. Onal E, Lopata M, Ginzburg AS, OConnor TD. Diaphragmatic EMG
and transdiaphragmatic pressure measurements with a single catheter. Am Rev Respir Dis 1981;124:563565.
10. McKenzie DK, Gandevia SC. Phrenic nerve conduction times and twitch
pressures of the human diaphragm. J Appl Physiol 1985;58:14961504.
11. Javaheri S, Vinegar A, Smith J, Donovan E. Use of a modified SwanGanz pacing catheter for measuring Pdi and diaphragmatic EMG.
Pflugers Arch 1987;408:642645.
12. Yernault J. Lung mechanics. I. Lung elasticity. In: Quanjer PH, editor.
Standardized lung function testing, Chapter 4. Bull Eur Physiopathol
Respir 1983;19(Suppl 5):2832.
13. Agostoni E, Rahn H. Abdominal and thoracic pressures at different
lung volumes. J Appl Physiol 1960;15:10871092.
14. Baydur, A., Behrakis PK, Zin WA, Jaeger M, Milic-Emili J. A simple
method for assessing the validity of the esophageal balloon technique. Am Rev Respir Dis 1982;126:788791.
15. Jacobs R, Killam H, Barefoot C, Millar H. Human application of a
catheter with tip-mounted pressure and flow transducers. Rev Surg
1972;29:149152.
16. Millar HD, Baker LE. A stable ultraminiature catheter-tip pressure
transducer. Med Biol Eng 1973;11:8689.
17. Gilbert R, Peppi D, Auchincloss JH Jr. Measurement of transdiaphragmatic pressure with a single gastric-esophageal probe. J Appl Physiol
1979;47:628630.
18. Evans SA, Watson L, Cowley AJ, Johnston ID, Kinnear WJ. Normal
range for transdiaphragmatic pressures during sniffs with catheter
mounted transducers. Thorax 1993;48:750753.
19. Wald A, Post K, Ransohoff J, Hass W, Epstein F. A new technique for monitoring epidural intracranial pressure. Med Instrum 1977;11:352354.
20. Yellowlees IH. Fibreoptic sensors in clinical measurement. Br J Anaesth 1991;67:100105.
21. Shapiro S, Bowman R, Callahan J, Wolfla C. The fiberoptic intraparenchymal cerebral pressure monitor in 244 patients. Surg Neurol 1996;
45:278282.
22. Koska J, Kelley E, Banner MJ, Blanch P. Evaluation of a fiberoptic system for airway pressure monitoring. J Clin Monit 1994;10:247250.
23. Heritier F, Rahm F, Pasche P, Fitting JW. Sniff nasal inspiratory pressure: a noninvasive assessment of inspiratory muscle strength. Am J
Respir Crit Care Med 1994;150:16781683.
24. Mead J. Mechanical properties of the lung. Physiol Rev 1961;41:281330.
25. De Troyer A, Estenne M. Limitations of measurement of transdia-
544
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.

phragmatic pressure in detecting diaphragmatic weakness. Thorax
1981;36:169174.
Black L, Hyatt R. Maximal respiratory pressures: normal values and relationship to age and sex. Am Rev Respir Dis 1969;99:696702.
Koulouris N, Mulvey DA, Laroche CM, Green M, Moxham J. Comparison of two different mouthpieces for the measurement of PImax and
PEmax in normal and weak subjects. Eur Respir J 1988;1:863867.
Hamnegard CH, Wragg S, Kyroussis D, Aquilina R, Moxham J, Green
M. Portable measurement of maximum mouth pressures. Eur Respir
J 1994;7:398401.
Gandevia SC, McKenzie DK. Activation of the human diaphragm during maximal static efforts. J Physiol 1985;367:4556.
Bigland-Ritchie BR, Furbush FH, Gandevia SC, Thomas CK. Voluntary discharge frequencies of human motor neurones at different
muscle lengths. Muscle Nerve 1992;15:130137.
Allen GM, Gandevia SC, McKenzie DK. Reliability of measurements
of muscle strength and voluntary activation using twitch interpolation. Muscle Nerve 1995;18:593600.
De Troyer A, Legrand A, Gevenois P-A, Wilson TA. Mechanical advantage of the human parasternal intercostal and triangularis sterni
muscles. J Physiol 1998;513:915925.
Agostoni E, Mead J. Statics of the respiratory system. In: Fenn WO
and Rahn H, editors. Handbook of physiology: respiration, Vol. 1,
Section 3. Washington DC: American Physiology Society; 1964. p.
387409.
Farkas GA, Rochester DF. Functional characteristics of canine costal
and crural diaphragm. J Appl Physiol 1988;65:22532260.
Road J, Newman S, Derenne JP, Grassino A. In vivo lengthforce relationship of canine diaphragm. J Appl Physiol 1986;60:6370.
Braun NMT, Arora NS, Rochester DF. Forcelength relationship of
the normal human diaphragm. J Appl Physiol 1982;53:405412.
DeTroyer A, Borenstein S, Cordier R. Analysis of lung volume restriction in
patients with respiratory muscle weakness. Thorax 1980;35:603610.
Black LF, Hyatt RE. Maximal static respiratory pressures in generalised neuromuscular disease. Am Rev Respir Dis 1971;103:641650.
Rochester DF. Tests of respiratory muscle function. Clin Chest Med
1988;9:249261.
Rinqvist T. The ventilatory capacity in healthy subjects: an analysis of
causal factors with special reference to the respiratory forces. Scand
J Clin Lab Invest 1966;18:8170.
Rochester DF, Arora NS. Respiratory muscle failure. Med Clin North
Am 1983;67:573598.
Leech JA, Ghezzo H, Stevens D, Becklake MR. Respiratory pressures
and function in young adults. Am Rev Respir Dis 1983;128:1723.
Wilson SH, Cooke NT, Edwards RHT, Spiro SG. Predicted normal values for maximal respiratory pressures in Caucasian adults and children. Thorax 1984;39:535538.
Vincken W, Ghezzo H, Cosio MG. Maximal static respiratory pressure
in adults: normal values and their relationship to determinants of respiratory function. Bull Eur Physiopathol Respir 1987;23:435439.
Aldrich TK, Spiro P. Maximal inspiratory pressure: does reproducibility indicate full effort? Thorax 1995;50:4043.
Enright PL, Kronmal RA, Monolio TA, Schenker MB, Hyatt RE. Respiratory muscle strength in the elderly. Am J Respir Crit Care Med
1994;149:430438.
McElvany G, Blackie S, Morrison NJ, Wilcox GP, Fairbarn MS, Pardy
RL. Maximal static respiratory pressures in the normal elderly. Am
Rev Respir Dis 1989;139:277281.
Enright PL, Kronmal RA, Higgins M, Schenker M, Haponik EF.
Spirometry reference values for women and men ages 6585. Cardiovascular Health Study. Am Rev Respir Dis 1993;147:125133.
Gaultier C, Zinman R. Maximal static pressures in healthy children.
Respir Physiol 1983;51:4561.
Wagener JS, Hibbert ME, Landau LI. Maximal respiratory pressures in
children. Am Rev Respir Dis 1984;129:873875.
Cook CD, Mead J, Orzalesi M. Static volumepressure characteristics
of the respiratory system during maximal efforts. J Appl Physiol
1964;19:10161022.
Hershenon MA, Kikuchi Y, Loring SH. Relative strengths of the chest
wall muscles. J Appl Physiol 1988;65:852862.
Laporta D, Grassino A. Assessment of transdiaphragmatic pressure in
humans. J Appl Physiol 1985;58:14691476.
Hillman DR, Markos J, Finucane KE. Effect of abdominal compression on maximum transdiaphragmatic pressure. J Appl Physiol 1990;
68:22962304.
Mier-Jedrzejowicz A, Brophy C, Moxham J, Green M. Assessment of
diaphragm weakness. Am Rev Respir Dis 1988;137:877883.
VOL 166
2002
56. Bellemare F, Bigland-Ritchie B. Assessment of human diaphragm

strength and activation using phrenic nerve stimulation. Respir Physiol 1984;58:263277.
57. Alexander C. Diaphragm movements and the diagnosis of diaphragmatic paralysis. Clin Radiol 1966;17:7983.
58. Hitzenberger K. Das Zwerchfell in gedunden und beranben Zustand.
Vienna, Austria: Springer; 1927.
59. Esau SA, Bye PT, Pardy RL. Changes in rate of relaxation of sniffs with
diaphragmatic fatigue in humans. J Appl Physiol 1983;55:731735.
60. Esau SA, Bellemare F, Grassino A, Permutt S, Roussos C, Pardy RL.
Changes in relaxation rate with diaphragmatic fatigue in humans. J
Appl Physiol 1983;54:13531360.
61. Miller JM, Moxham J, Green M. The maximal sniff in the assessment of
diaphragm function in man. Clin Sci 1985;69:9196.
62. Laroche CM, Mier AK, Moxham J, Green M. The value of sniff esophageal pressures in the assessment of global inspiratory muscle strength.
63. Pertuze J, Watson A, Pride N. Limitation of maximum inspiratory flow
through the mouth [abstract]. Clin Respir Physiol 1987;23:34S.
64. Heritier F, Rahm F, Pasche P, Fitting J-W. Sniff nasal pressure: a noninvasive assessment of inspiratory muscle strength. Am J Respir Crit
Care Med 1994;150:16781683.
65. Koulouris N, Vianna LG, Mulvey DA, Green M, Moxham J. Maximal
relaxation rates of esophageal, nose, and mouth pressures during a
sniff reflect inspiratory muscle fatigue. Am Rev Respir Dis 1989;139:
12131217.
66. Uldry C, Fitting J. Influence of airway obstruction on sniff nasal inspiratory pressure. Am J Respir Crit Care Med 1995;151:A414.
67. Uldry C, Janssens JP, de Meralt B, Fitting JW. Sniff nasal inspiratory
pressure in patients with chronic obstructive pulmonary disease. Eur
Respir J 1997;10:12921296.
68. Polkey MI, Green M, Moxham J. Measurement of respiratory muscle
strength. Thorax 1995;50:11311135.
69. Gandevia SC, Gorman RB, McKenzie DK, Southon FC. Dynamic
changes in human diaphragm length: maximal inspiratory and expulsive
efforts studied with sequential radiography. J Physiol 1992;457:167176.
70. Bouros D, Siafakas N, Green M. Cough: physiological and pathophysiological considerations. In: Roussos C, editor. The thorax. New York:
Marcel Dekker; 1995. p. 13351354.
71. Rahn H, Otis AB, Chadwick LE, Fenn WO. The pressurevolume diagram of the thorax and lung. Am J Physiol 1946;146:161178.
72. Krietzer S, Saunders M, Tyler HR, Ingram RH. Respiratory muscle
function in amyotrophic lateral sclerosis. Am Rev Respir Dis 1978;
117:437447.
73. Polkey MI, Lyall RA, Green M, Leigh PN, Moxham J. Expiratory muscle function in amyotrophic lateral sclerosis. Am J Respir Crit Care
Med 1998;158:734741.
74. Coryllos PN. Action of the diaphragm in cough. Am J Med Sci 1937;
194:523535.
75. De Troyer A, Sampson M, Sigrist S, Kelly S. How the abdominal muscles act on the rib cage. J Appl Physiol 1983;58:14381443.
76. Kyroussis D, Polkey MI, Hughes PD, Fleming TA, Wood CN, Mills
GH, Hamnegard C-H, Green M, Moxham J. Abdominal muscle
strength measured by gastric pressure during maximal cough. Thorax
1996;51(Suppl 3):A45.
77. Cox ID, Osman RCA, Hughes DTD, Empey DW. The abdominal electromyogram as an objective measure of cough intensity. Thorax 1983;
38:222.
78. Black LF, Hyatt RE. Maximal respiratory pressures: normal values and
relationships to age and sex. Am Rev Respir Dis 1969;99:696702.
79. Loudon RG, Shaw GB. Mechanics of cough in normal subjects and in
patients with obstructive respiratory disease. Am Rev Respir Dis
1967;96:666677.
80. National Heart, Lung, and Blood Institute. Respiratory muscle fatigue.
NHLBI Workshop. Am Rev Respir Dis 1990;142:474480.
81. Aubier M, Farkas G, De Troyer A, Mozes R, Roussos C. Detection of
diaphragmatic fatigue in man by phrenic stimulation. J Appl Physiol
1981;50:538544.
82. Mier A, Brophy C. Measurement of twitch transdiaphragmatic pressure:
surface versus needle electrode stimulation. Thorax 1991;46:669670.
83. Hubmayr RD, Litchy WJ, Gay PC, Nelson SB. Transdiaphragmatic
twitch pressure: effects of lung volume and chest wall shape. Am Rev
Respir Dis 1989;139:647652
84. Mier A, Brophy C, Moxham J, Green M. Influence of lung volume and
rib cage configuration on transdiaphragmatic pressure during phrenic
nerve stimulation in man. Respir Physiol 1990;80:193202.
85. Eastwood PR, Panizza JA, Hillman DR, Finucane KE. Application of
545
86.
87.
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
a cervical stimulating apparatus for bilateral transcutaneous phrenic

nerve stimulation. J Appl Physiol 1995;79:632637.
Koulouris N, Mulvey DA, Laroche CM, Goldstone J, Moxham J,
Green M. The effect of posture and abdominal binding on respiratory pressures. Eur Respir J 1989;2:961965.
Wilcox PG, Eisen A, Wiggs BJ, Pardy RL. Diaphragmatic relaxation
rate after voluntary contractions and uni- and bilateral phrenic stimulation. J Appl Physiol 1988;65:675682.
Bellemare F, Bigland-Ritchie B. Central components of diaphragmatic
fatigue assessed by phrenic nerve stimulation. J Appl Physiol 1987;
62:13071316.
Yan S, Gauthier AP, Similowski T, Faltus R, Macklem PT, Bellemare
F. Forcefrequency relationships of in vivo human and in vitro rat diaphragm using paired stimuli. Eur Respir J 1993;6:211218.
Mador MJ, Magalang UJ, Rodis A, Kufel TJ. Diaphragmatic fatigue after exercise in healthy human subjects. Am Rev Respir Dis 1993;148:
15711575.
Desmedt JE, Hainaut K. Kinetics of myofilament activation in potentiated contraction: staircase phenomenon in human skeletal muscle.
Nature 1968;217:529532.
Van Lunteren E, Vafaie H. Force potentiation in respiratory muscles:
comparison of diaphragm and sternohyoid. Am J Physiol 1993;264:
R1095R1100.
DArsonval A. Production des courants de haute frquence et de
grande intensit; leurs effets physiologiques. C R Soc Biol 1893;45:
122124.
Olney RK, So YT, Goodin DS, Aminoff MJ. A comparison of magnetic and electrical stimulation of peripheral nerves. Muscle Nerve
1990;13:957963.
Maccabee PJ, Amassian VE, Cracco RQ, Eberle LP, Rudell AP. Mechanisms of peripheral nervous system stimulation using the magnetic
coil. Electroencephalogr Clin Neurophysiol Suppl 1991;43:344361.
Schmid UD, Walker G, Schmid-Sigron J, Hess CW. Transcutaneous
magnetic and electrical stimulation over the cervical spine: excitation
of plexus roots rather than spinal roots. Electroencephalogr Clin
Neurophysiol Suppl 1991;43:369384.
Ono S, Oishi M, Du CM, Takasu T. Magnetic stimulation of peripheral
nerves: comparison of magnetic stimulation with electrical stimulation. Electromyogr Clin Neurophysiol 1995;35:317320.
Chokroverty S. Magnetic stimulation in clinical neurophysiology. London: Butterworth; 1990.
Barker AT, Jalinous R, Freeston IL. Noninvasive magnetic stimulation
of human motor cortex. Lancet 1985;8437:11061107.
Similowski T, Fleury B, Launois S, Cathala HP, Bouche P, Derenne JP.
Cervical magnetic stimulation: a new painless method for bilateral
phrenic nerve stimulation in conscious humans. J Appl Physiol 1989;
67:13111318.
Chokroverty S, Shah S, Chokroverty M, Deutsch A, Belsh J. Percutaneous magnetic coil stimulation of the phrenic nerve roots and trunk.
Electroencephalogr Clin Neurophysiol 1995;97:369374.
Chen R, Collins S, Remtulla H, Parkes A, Bolton CF. Phrenic nerve
conduction study in normal subjects. Muscle Nerve 1995;18:330335.
Zifko U, Remtulla H, Power K, Harker L, Bolton CF. Transcortical
and cervical magnetic stimulation with recording of the diaphragm.
Muscle Nerve 1996;19:614620.
Maccabee PJ, Amassian VE, Eberle LP, Rudell AP, Cracco RQ, Lai
KS, Somasundarum M. Measurement of the electric field induced
into inhomogeneous volume conductors by magnetic coils: application to human spinal neurogeometry. Electroencephalogr Clin Neurophysiol 1991;81:224237.
Mills KR, McLeod C, Sheffy J, Loh L. The optimal current direction
for excitation of human cervical motor roots with a double coil
magnetic stimulator. Electroencephalogr Clin Neurophysiol 1993;89:
138144.
Similowski T, Mehiri S, Attali V, Duguet A, Straus C, Derenne J-P.
Comparison of magnetic and electrical phrenic nerve stimulation in
assessment of phrenic nerve conduction time. J Appl Physiol 1997;
82:11901199.
Wragg S, Aquilina R, Moran J, Ridding M, Hamnegard C, Fearn T,
Green M, Moxham J. Comparison of cervical magnetic stimulation
and bilateral percutaneous electrical stimulation of the phrenic nerves
in normal subjects. Eur Respir J 1994;7:17881792.
Laghi F, Harrison MJ, Tobin MJ. Comparison of magnetic and electrical phrenic nerve stimulation in assessment of diaphragmatic contractility. J Appl Physiol 1996;80:17311742.
Similowski T, Duguet A, Straus C, Boisteanu D, Attali V, Derenne J-P.
Assessment of the voluntary activation of the diaphragm in man us-
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
ing cervical and cortical magnetic stimulation. Eur Respir J 1996;9:

12241231.
Hamnegard CH, Wragg S, Kyroussis D, Mills G, Bake B, Green M,
Moxham J. Mouth pressure in response to magnetic stimulation of
the phrenic nerves. Thorax 1995;50:620624.
Rajanna MJ. Anatomical and surgical considerations of the phrenic
and accessory phrenic nerves. J Int Coll Surgeons 1947;60:4253.
Sarnoff SJ, Sarnoff LC, Whittenberger JL. Electrophrenic respiration.
VII. The motor point of the phrenic nerve in relation to external
stimulation. Surg Gynecol Obstet 1951;93:190196.
Luo YM, Polkey MI, Johnson LC, Lyall RA, Harris ML, Green M, Moxham J. Diaphragm EMG measured by cervical magnetic and electrical
phrenic nerve stimulation. J Appl Physiol 1998;85:20892099.
Mills GH, Kyroussis D, Hamnegard CH, Wragg S, Moxham J, Green
M. Unilateral magnetic stimulation of the phrenic nerve. Thorax
1995;50:11621172.
Mills G, Kyroussis D, Hamnegard C, Polkey M, Green M, Moxham J.
Bilateral magnetic stimulation of the phrenic nerves from an anterolateral approach. Am J Respir Crit Care Med 1996;154:10991105.
Bellemare F, Bigland-Ritchie B, Woods JJ. Contractile properties of
the human diaphragm in vivo. J Appl Physiol 1986;61:11531161.
Petitjean M, Ripart J, Couture J, Bellemare F. Effects of lung volume
and fatigue on evoked diaphragmatic phonomyogram in normal subjects. Thorax 1996;51:705710.
Aubier M, Murciano D, Lecocguic Y, Viires N, Pariente R. Bilateral
phrenic stimulation: a simple technique to assess diaphragmatic fatigue in humans. J Appl Physiol 1985;58:5864.
Yan S, Gauthier AP, Similowski T, Macklem PT, Bellemare F. Evaluation of human diaphragm contractility using mouth pressure twitches.
Am Rev Respir Dis 1992;145:10641069.
Similowski T, Gauthier AP, Yan S, Macklem PT, Bellemare F. Assessment of diaphragm function using mouth pressure twitches in
chronic obstructive pulmonary disease patients. Am Rev Respir Dis
1993;147:850856.
Hughes PD, Polkey MI, Kyroussis D, Hamnegard C-H, Moxham J,
Green M. Measurement of sniff nasal and diaphragm twitch mouth
pressure in patients. Thorax 1998;53:96100.
Rahn H, Otis AB, Chadwick L, Fenn O. The pressure volume diagram
of the thorax and lung. Am J Physiol 1946;146:161178.
Byrd RB, Hyatt RE. Maximal respiratory pressures in chronic obstructive lung disease. Am Rev Respir Dis 1968;98:848856.
Decramer M, Demedts M, Rochette F, Billiet L. Maximal transrespiratory pressures in obstructive lung disease. Bull Eur Physiopathol
Respir 1980;16:479490.
Loring SH, Mead J, Griscom NT. Dependence of diaphragmatic length
on lung volume and thoracoabdominal configuration. J Appl Physiol
1985;59:19611970.
Heijdra YF, Dekhuijzen PN, van Herwaarden CL, Folgering HT. Effects of body position, hyperinflation, and blood gas tensions on
maximal respiratory pressures in patients with chronic obstructive
pulmonary disease. Thorax 1994;49:453458.
Wanke T, Merkle M, Zifko U, Formanek D, Lahrmann H, Grisold W,
Zwick H. The effect of aminophylline on the forcelength characteristics of the diaphragm. Am J Respir Crit Care Med 1994;149:1545
1549.
Tolep K, Higgins N, Muza S, Criner G, Kelsen SG. Comparison of diaphragm strength between healthy adult elderly and young men. Am J
Hamnegard CH, Wragg S, Mills G, Kyroussis D, Road J, Daskos G,
Bake B, Moxham J, Green M. The effect of lung volume on transdiaphragmatic pressure. Eur Respir J 1995;8:15321536.
Smith J, Bellemare F. Effect of lung volume on in vivo contraction characteristics of human diaphragm. J Appl Physiol 1987;62:18931900.
Similowski T, Yan S, Gauthier AP, Macklem PT, Bellemare F. Contractile properties of the human diaphragm during chronic hyperinflation. N Engl J Med 1991;325:917923.
Yan S, Similowski T, Gauthier AP, Macklem PT, Bellemare F. Effect
of fatigue on diaphragmatic function at different lung volumes. J
Appl Physiol 1992;72:10641067.
Gauthier AP, Yan S, Sliwinski P, Macklem PT. Effects of fatigue, fiber
length, and aminophylline on human diaphragm contractility. Am J
Polkey MI, Kyroussis D, Keilty SE, Hamnegard CH, Mills GH, Green
M, Moxham J. Exhaustive treadmill exercise does not reduce twitch
transdiaphragmatic pressure in patients with COPD. Am J Respir
Crit Care Med 1995;152:959964.
Gauthier AP, Verbanck S, Estenne M, Segebarth C, Macklem PT, Paiva
546
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.

M. Three-dimensional reconstruction of the in vivo human diaphragm
shape at different lung volumes. J Appl Physiol 1994;76:495506.
Loring SH. Three-dimensional reconstruction of the in vivo human diaphragm shape at different lung volumes [editorial]. J Appl Physiol
1994;76:493494.
Tabary JC, Tabary C, Tardieu C, Tardieu G, Goldspink G. Physiological and structural changes in the cats soleus muscle due to immobilization at different lengths by plaster casts. J Physiol 1972;224:231244.
Farkas GA, Roussos C. Diaphragm in emphysematous hamsters: sarcomere adaptability. J Appl Physiol 1983;54:16351640.
Farkas GA, Roussos C. Adaptability of the hamster diaphragm to exercise and/or emphysema. J Appl Physiol 1983;53:12631272.
Oliven A, Supinski G, Kelsen SG. Functional adaptation of diaphragm
to chronic hyperinflation in emphysematous hamsters. J Appl Physiol 1986;60:225231.
Matano T, Tamai K, Kurokawa T. Adaptation of skeletal muscle in
limb lengthening: a light diffraction study on the sarcomere length in
situ. J Orthop Res 1994;12:193196.
Prezant DJ, Aldrich TK, Karpel JP, Lynn RI. Adaptations in the diaphragms in vitro forcelength relationship in patients on continuous
ambulatory peritoneal dialysis. Am Rev Respir Dis 1990;141:13421349.
Wilson DF, Skirboll LR. Basis for posttetanic potentiation at the mammalian neuromuscular junction. Am J Physiol 1974;227:9295.
Vandervoort AA, Quinlan J, McComas AJ. Twitch potentiation after
voluntary contraction. Exp Neurol 1983;81:141152.
Mador MJ, Magalang UJ, Kufel TJ. Twitch potentiation following voluntary diaphragmatic contraction. Am J Respir Crit Care Med 1994;
149:739743.
Wragg S, Hamnegard C, Road J, Kyroussis D, Moran J, Green M,
Moxham J. Potentiation of diaphragmatic twitch after voluntary contraction in normal subjects. Thorax 1994;49:12341237.
Desmedt JE, Hainaut K. Modifications des proprits contractiles du
muscle stri au cours de la stimulation lectrique rpte de son nerf
moteur chez lhomme normal. C R Acad Sci Paris 1967;264:363366.
Mills G, Kyroussis D, Hamnegard C, Wragg S, Moxham J, Green M.
Cervical magnetic stimulation of the phrenic nerves in bilateral diaphragm paralysis. Am J Respir Crit Care Med 1997;155:15651569.
Attali V, Mehiri S, Straus C, Salachas F, Meininger V, Derenne J-P,
Similowski T. Influence of neck muscles hypertrophy on mouth pressure response to cervical magnetic stimulation [abstract]. Am J
Respir Crit Care Med 1996;153:A786.
Merton PA. Voluntary strength and fatigue. J Physiol 1954;67:553564.
Bellanger AY, McComas AJ. Extent of motor unit activation during effort. J Appl Physiol 1981;51:11311135.
Behm DG, St-Pierre MM, Perez D. Muscle inactivation: assessment of
interpolated twitch technique. J Appl Physiol 1996;81:22672273.
Rutherford OM, Jones DA, Newham DJ. Clinical and experimental
application of the percutaneous twitch superimposition technique
for the study of human muscle activation. J Neurol Neurosurg Psychiatr 1986;49:12881291.
Gandevia SC, McKenzie DK. Human diaphragmatic endurance during
different maximal respiratory efforts. J Physiol 1988;395:625638.
Gandevia SC, McKenzie DK, Plassman BL. Activation of human respiratory muscles during different voluntary manoeuvres. J Physiol 1990;
428:387403.
Loring SH, Hershensen MB. Effects of series compliance on twitches superimposed on voluntary contractions. J Appl Physiol 1992;73: 516521.
Close RI. Dynamic properties of mammalian skeletal muscles. Physiol
Rev 1972;52:129197.
Ferguson GT. Use of twitch pressures to assess diaphragmatic function
and central drive. J Appl Physiol 1994;77:17051715.
Bigland-Ritchie B, Jones DA, Hosking GP, Edwards RHT. Central
and peripheral fatigue in sustained maximum voluntary contractions
of human quadriceps muscle. Clin Sci Mol Med 1978;54:609614.
McKenzie DK, Bigland-Ritchie B, Gorman RB, Gandevia SC. Central
and peripheral fatigue of human diaphragm and limb muscles assessed by twitch interpolation. J Physiol 1992;454:643656.
Allen GM, Hickie I, Gandevia SC, McKenzie DK. Impaired voluntary
drive to breathe: a possible link between depression and unexplained
ventilatory failure in asthmatic patients. Thorax 1994;49:881884.
Loureno RV, Miranda JM. Drive and performance of the ventilatory
apparatus in chronic obstructive lung disease. N Engl J Med 1968;
279:5359.
McKenzie DK, Gandevia SC. Phrenic nerve conduction times and twitch
Hershenson MB, Kikuchi Y, Loring SH. Relative strengths of the chest
wall muscles. J Appl Physiol 1988;65:852862.
VOL 166
2002
165. McKenzie DK, Plassman BL, Gandevia SC. Maximal activation of the
human diaphragm but not inspiratory intercostal muscles during
static inspiratory efforts. Neurosci Lett 1988;89:6368.
166. Levy RD, Nava S, Gibbons L, Bellemare F. Aminophylline and human
diaphragm strength in vivo. J Appl Physiol 1990;68:25912596.
167. Allen GM, McKenzie DK, Gandevia SC, Bass S. Reduced voluntary
drive to breathe in asthmatic subjects. Respir Physiol 1993;93:2940.
168. Moxham J, Morris AJ, Spiro SG, Edwards RH, Green M. Contractile
properties and fatigue of the diaphragm in man. Thorax 1981;36:
164168.
169. Bai TR, Rabinovitch BJ, Pardy RL. Near-maximal voluntary hyperpnea
and ventilatory muscle function. J Appl Physiol 1984;57:17421748.
170. Johnson BD, Babcock MA, Suman OE, Dempsey JA. Exercise-induced
diaphragmatic fatigue in healthy humans. J Physiol 1993;460:385405.
171. Babcock MA, Johnson BD, Pegelow DF, Suman OE, Griffin D, Dempsey JA. Hypoxic effects on exercise-induced diaphragmatic fatigue in
normal healthy humans. J Appl Physiol 1995;78:8292.
172. Babcock MA, Pegelow DF, McClaran SR, Suman OE, Dempsey JA.
Contribution of diaphragmatic power output to exercise-induced diaphragm fatigue. J Appl Physiol 1995;78:17101719.
173. Yan S, Gauthier AP, Similowski T, Faltus R, Macklem PT, Bellemare
F. Forcefrequency relationships of in vivo human and in vitro rat diaphragm using paired stimuli. Eur Respir J 1993;6:211218.
174. Polkey MI, Kyroussis D, Mills GH, Hughes PD, Moxham J, Green M.
Paired phrenic nerve stimuli for the detection of diaphragm fatigue
in humans. Eur Respir J 1997;10:18591864.
175. Hughes PD, Polkey MI, Harrus ML, Coats AJ, Moxham J, Green M.
Diaphragm strength in chronic heart failure. Am J Respir Crit Care
Med 1999;160:529534.
176. Dureuil B, Viires N, Cantineau JP, Aubier M, Desmonts JM. Diaphragmatic contractility after upper abdominal surgery. J Appl Physiol 1986;61:17751780.
177. Derrington MC, Hindocha N. Measurement of evoked diaphragm
twitch strength during anaesthesia: adaptation and evaluation of an
existing technique. Br J Anaesth 1988;61:270278.
178. Similowski T, Fleury B, Launois S, Cathala HP, Bouche P, Derenne JP.
Stimulation magntique cervicale (SMC). Une nouvelle mthode de
stimulation phrnique bilatrale utilisable en clinique. Rev Mal Respir
1988;5:609614.
179. Mier A, Brophy C, Moxham J, Green M. Twitch pressures in the assessment of diaphragm weakness. Thorax 1989;44:990996.
180. Geddes LA, Mouchawar G, Bourland JD, Nyenhuis J. Inspiration produced by bilateral electromagnetic, cervical phrenic nerve stimulation in man. IEEE Trans Biomed Eng 1991;38:10471048.
181. McKenzie DK, Gandevia SC. Recovery from fatigue of human diaphragm and limb muscles. Respir Physiol 1991;84:4960.
182. Mancini DM, Henson D, LaManca J, Levine S. Respiratory muscle
function and dyspnea in patients with chronic congestive heart failure. Circulation 1992;86:909918.
183. Wanke T, Merkle M, Formanek D, Zifko U, Wieselthaler G, Zwick H,
Klepetko W, Burghuber OC. Effect of lung transplantation on diaphragmatic function in patients with chronic obstructive pulmonary
disease. Thorax 1994;49:459464.
184. Laghi F, DAlfonso N, Tobin MJ. Pattern of recovery from diaphragmatic fatigue over 24 hours. J Appl Physiol 1995;79:539546.
185. Babcock MA, Pegelow DF, Johnson BD, Dempsey JA. Aerobic fitness
effects on exercise-induced low-frequency diaphragm fatigue. J Appl
Physiol 1996;81:21562164.
186. Hamnegard CH, Wragg S, Kyroussis D, Mills GH, Polkey MI, Moran J,
Road J, Bake B, Green M, Moxham J. Diaphragm fatigue following
maximal ventilation in man. Eur Respir J 1996;9:241247.
187. Hamnegard CH, Wragg SD, Mills GH, Polkey MI, Bake B, Moxham J,
Green M. Clinical assessment of diaphragm strength by cervical magnetic stimulation of the phrenic nerves. Thorax 1996;51:12391242.
188. Mador JM, Rodis A, Diaz J. Diaphragmatic fatigue following voluntary
hyperpnea. Am J Respir Crit Care Med 1996;154:6367.
189. Mador M, Dahuja M. Mechanisms for diaphragmatic fatigue following highintensity leg exercise. Am J Respir Crit Care Med 1996;154:14841489.
190. Polkey M, Kyroussis D, Hamnegard C-H, Mills G, Green M, Moxham
J. Diaphragm strength in chronic obstructive pulmonary disease. Am
J Respir Crit Care Med 1996;154:13101317.
191. Aubier M, Murciano D, Lecocguic Y, Viires N, Jacquens Y, Squara P,
Pariente R. Effect of hypophosphatemia on diaphragmatic contractility in patients with acute respiratory failure. N Engl J Med 1985;
313:420424.
192. Aubier M, Murciano D, Viires N, Lebargy F, Curran Y, Seta JP, Pariente R. Effects of digoxin on diaphragmatic strength generation in
547
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
203.
patients with chronic obstructive pulmonary disease during acute respiratory failure. Am Rev Respir Dis 1987;135:544548.
Laroche CM, Cairns T, Moxham J, Green M. Hypothyroidism presenting with respiratory muscle weakness. Am Rev Respir Dis 1988;138:
472474.
Laroche CM, Carroll N, Moxham J, Green M. Clinical significance of
severe isolated diaphragm weakness. Am Rev Respir Dis 1988;138:
862866.
Aubier M, Murciano D, Menu Y, Boczkowski J, Mal H, Pariente R.
Dopamine effects on diaphragmatic strength during acute respiratory failure in chronic obstructive pulmonary disease. Ann Intern
Med 1989;110:1723.
Murciano D, Rigaud D, Pingleton S, Armengaud MH, Melchior JC,
Aubier M. Diaphragmatic function in severely malnourished patients with anorexia nervosa. Effects of renutrition. Am J Respir Crit
Care Med 1994;150:15691574.
Hsiun-ing C, Kuo CS. Relationhip between respiratory muscle function
and age, sex, and other factors. J Appl Physiol 1989;66:943948.
Similowski T, Straus C, Attali V, Duguet A, Jourdain B, Derenne J-P.
Assessment of the motor pathway to the diaphragm using cortical
and cervical magnetic stimulation in the decision making process of
phrenic pacing. Chest 1996;110:15511557.
Nochomovitz ML, Hopkins M, Brodkey J, Montenegro H, Mortimer
JT, Cherniack NS. Conditioning of the diaphragm with phrenic
nerve stimulation after prolonged disuse. Am Rev Respir Dis 1984;
130:685688.
Mier A, Brophy C, Estenne M, Moxham J, Green M, De Troyer A. Action of abdominal muscles on rib cage in humans. J Appl Physiol
1985;58:14381443.
Linder SH. Functional electrical stimulation to enhance cough in quadriplegia. Chest 1993;103:166169.
Kyroussis D, Polkey MI, Mills GH, Hughes PD, Moxham J, Green M.
Stimulation of cough in man by magnetic stimulation of the thoracic
nerve roots. Am J Respir Crit Care Med 1997;156:16961699.
Kyroussis D, Mills GH, Polkey MI, Hamnegaard C-H, Koulouris N,
Green M, Moxham J. Abdominal muscle fatigue after maximal ventilation in humans. J Appl Physiol 1996;81:14771483.
APPENDIX: ABBREVIATIONS
Measures
P
L
V
EMG
PMG
Pressure (usually mouth pressure if site not specified, as PImax)

Length
Volume
Electromyogram
Phonomyogram
Sites and Modifiers

A,
alv
ab
abw
ao
aw
bs
di
es, oes
g, ga
Alveolar
Abdomen
Abdominal wall
Airway opening
Airway
Body surface
Diaphragm, transdiaphragmatic
Esophageal
Gastric
ia
mo
mus
nas
np
ph
pl
rc
rs
w, cw
ant
post
E
I
Intercostal/accessory muscles
Mouth
Muscle
Nostril, nasal
Nasopharynx
Phrenic
Pleural
Rib cage
Respiratory system (lung and chest wall)
Chest wall
Anterior
Posterior
Expiratory
Inspiratory
Maneuvers
co
max
sn
tw
vol
Cough
Maximal
Sniff
Twitch
Voluntary
Stimulation Descriptors
ELS
MS
Ant MS
CMS
fMS
PNS
EMG
PMG
M-wave
Electrical stimulation
Magnetic stimulation
Anterior magnetic stimulation
Cervical magnetic stimulation
Focal magnetic stimulation
Phrenic nerve stimulation
Electromyogram
Phonomyogram
EMG response to PNS
Lung Volumes
FRC
RV
TLC
VC
Functional residual capacity

Residual volume
Total lung capacity
Vital capacity
General
CNS
HFF
LFF
MVC
SMVC
Central nervous system

High-frequency fatigue
Low-frequency fatigue
Maximal voluntary contractions
Submaximal voluntary contractions
Principles
Give measure, then site, then maneuver, then descriptor. Thus:

Pes,tw,CMS twitch esophageal pressure following cervical
magnetic stimulation
PE,mo,max maximal expiratory mouth pressure (usually abbreviated to PEmax)
Pg,co gastric pressure during a cough
3. Electrophysiologic Techniques for the Assessment of

Respiratory Muscle Function
Respiratory muscle contraction depends on electrical activation of the muscles. Influenced both by involuntary and voluntary inputs, the electrical impulses originate in the respiratory
neurons of the brainstem, are carried via motor nerves, transmit through neuromuscular junctions, and propagate throughout muscle membranes. Failure at any of these sites can result
in dyscoordination and reversible or irreversible muscle weakness. The task of electrophysiologic tests is to assess the integrity of the respiratory neuromotor apparatus.
There are two main types of electrophysiologic tests of respiratory muscle function: electromyography and stimulation
tests. These tests are interrelated, and they are related also to
tests of mechanical action of the respiratory muscles, described elsewhere in this Statement.
ELECTROMYOGRAPHY
Rationale
Electromyography (EMG) is the art of describing myoelectric

signals (1), the electrical manifestations of the excitation process elicited by action potentials propagating along muscle fiber membranes. The EMG signal is detected with electrodes,
and then amplified, filtered, and displayed on a screen or digitized to facilitate further analysis. Electromyography of respiratory muscles can be used to assess the level and pattern of
their activation, so as to detect and diagnose neuromuscular
pathology, and, when coupled with tests of mechanical function, to assess the efficacy of the muscles contractile function
(see ELECTROMECHANICAL EFFECTIVENESS in Section 6 of this
Statement).
Scientific Basis
Single fiber action potential. Depolarization of a muscle fiber

membrane, caused by the flow of ions across the sarcolemma,
generates an electric field outside the muscle fiber, which can
be detected by extracellular recording electrodes as voltage
changes over time; this voltage transient is known as the action
potential. Although the transmembrane potential changes generated by depolarization of a given fiber are always identical
in shape and amplitude (the all or nothing phenomenon),
the shape and amplitude of a recorded action potential depend on factors such as the orientation of the recording electrodes with respect to the active muscle fibers, the distance between the muscle fibers and the electrodes, the filtering
properties of the electrodes, and the muscle fiber action potential conduction velocity (2).
In humans, muscle fiber conduction velocity ranges from 2
to 6 m/second (3), depending on passive and active components of the muscle fiber membrane. The passive components
(cable properties) include capacitance per unit length (proportional to fiber circumference) and internal resistance (inversely proportional to the square of the fiber diameter). The
active components (membrane excitability) depend on ion
gradients across the membrane and properties of the ion gating channels, which in turn are influenced by electric field
strength, temperature, and chemical milieu (especially pH and
Ca2 concentrations). Muscle fiber conduction velocity has
been shown to vary with fiber diameter (4), temperature (3),
electrolyte
gradients across the cell membrane (5), pH (5),
DOI:
10.1164/rccm.AT0302
Internet
address:
and fatigue
(1,www.atsjournals.org
6).
Single motor unit action potential. Each motor unit is composed of a number of individual muscle fibers innervated by a
single anterior horn cell. All individual fibers within a motor
unit are activated almost simultaneously. The amplitude and
shape of the resulting motor unit action potential (MUAP)
are influenced not only by all the factors that can affect single
fiber action potentials, but also by such factors as the number
of fibers within the motor unit, the spatial dispersion of motor
unit fibers, differences in length of the motor neuron terminal
axons, and possibly fiber-to-fiber differences in action potential conduction velocity (2, 7).
Summation of motor unit signals. Compound muscle action
potentials (CMAPs) represent the summated electrical activity generated by all motor units synchronously activated by
nerve stimulation (see subsequent passages). The observed
CMAP is influenced by the number of activated motor units,
their synchronization, the shape of individual MUAPs, and
cancellation of opposite phase potentials.
The interference pattern EMG results from the temporal
and spatial summation of asynchronously firing MUAP trains
during spontaneous muscle contractions, when individual
MUAPs can no longer be distinguished (8). The observed interference pattern EMG is thus a function of the number of
active motor units, their firing rates and synchronization, the
shapes of their individual MUAPs (in turn dependent on all
the factors listed here previously), and cancellation of opposite phase potentials (2, 8).
EMG Equipment
Recording electrodes. Electromyography signals can be detected

as the difference between the signal from an electrode placed
on or in the muscle under investigation (active electrode) and
the signal from another electrode placed in an electrically silent region (indifferent or reference electrode). This electrode arrangement is usually referred to as a monopolar electrode. When the two electrodes connected to a differential
amplifier are positioned on or in the same muscle under investigation, the electrode arrangement is usually referred to as bipolar. Optimal EMG signals depend on the use of electrodes
with appropriate configuration and fixed geometry, on maintenance of electrode position relative to the muscle, on alignment of the electrodes with respect to fiber direction, on selection of sites with relatively low density of motor end plates,
and on avoidance of signal disturbances.
Electrodes can be placed on the skin overlying the neck,
the chest wall muscles, or the area of apposition of the diaphragm to the chest wall; they can be swallowed into the
esophagus to measure crural diaphragm EMG; or they can be
inserted into the respiratory muscle of interest, using needle,
wire, or hook electrodes. Selection of an appropriate electrode system requires consideration of advantages and disadvantages specific to the technique and the context of the study
(Table 1).
Surface electrodes. Surface electrodes have been used to
measure activity of diaphragm, intercostal, scalene, abdominal, and accessory muscles. After the skin is shaved, cleaned,
and dried, electrodes are placed over or as close as possible to
the muscle to be investigated and are secured with optimal
skin contact. The placement is determined by palpation and by
the investigators knowledge of respiratory muscle anatomy,
549

TABLE 1. TYPES OF RECORDING ELECTRODES FOR RESPIRATORY MUSCLE ELECTROMYOGRAMS
Type of Electrode
Surface electrodes
Chest wall
Esophageal
Intramuscular electrodes
Advantages
Disadvantages
Noninvasive
Large volume sampling
Cross-talk
Variable filtering
Less influenced by body habitus

Less cross-talk
Discomfort
Unreliable in diaphragm hernia
Less influenced by cross-talk

Single motor unit recordings possible
Discomfort
Difficult to place
Small pneumothorax risk
Possible sampling error
but no standards exist for electrode design or positioning. Furthermore, there is no consensus on methods either to maintain
electrode orientation with respect to muscle fibers and innervation zones or to control for influences of variable muscle-toelectrode distance (as, e.g., with variations in the amount of
subcutaneous fat), or for cross-talk from adjacent muscles.
Advantages of surface electrodes are their noninvasive nature and their ability to sample a large number of motor units.
For many individual nondiaphragm respiratory muscles, however, their proximity to one another and to nonrespiratory
trunk muscles makes surface electrode recordings unreliable.
Examples include cross-talk from scalenes and platysma in recordings of sternocleidomastoids (9), from external oblique
and pectoralis in recordings of intercostal muscles, and among
various abdominal muscles (10, 11). Furthermore, variations
in interindividual body habitus, for example, subcutaneous fat
tissue or deformity of the chest wall, produce variable muscleto-electrode filtering effects.
Esophageal electrodes. Esophageal electrodes are metal electrodes mounted on a catheter, which is inserted via the nose or
mouth and positioned with the electrode rings at the level of

the crural diaphragm. In adults, the motor innervation zone of
the crural diaphragm lies 13 cm cephalad to the gastroesophageal junction, with the left side approximately 1 cm cephalad
to the right (12).
Esophageal electrode catheters are often equipped with
gastric balloons and weights on the proximal end, which anchor them in position when outward traction pulls the gastric
balloon snugly against the gastroesophageal junction. This
feature may limit motion of the electrodes with respect to the
esophagus, but it does not prevent either diaphragm movement relative to the electrodes, or the resulting artifacts (13
16). Motion artifacts are minimized in semistatic contractions.
A more reliable method to reduce electrode positioning artifacts during dynamic maneuvers is continuous optimization
of diaphragm-electrode positioning. An electrode array of
eight rings mounted 1 cm apart on a catheter is introduced and
adjusted to provide optimal EMG activity from the central
pair of electrodes (17, 18) (Figure 1A). A computer program
samples all electrodes continuously, and selects the pair clos-
Figure 1. (A) Schematic description of the methods used

to locate the center of the electrically active region of the
diaphragm (EARdi,ctr). Left: Raw signals from each of
seven electrode pairs on an esophageal probe. Center: The
probe with its eight electrodes. Right: Correlation coefficients for cross-correlation of signals from various pairs of
electrodes. The dashed curve is the best fit curve, from
which the ERA-di ctr position is calculated. (B) An example
of the double subtraction technique. Diaphragm EMG signals from electrode pairs located 10 mm caudal and cephalad to the EAR-di ctr are shown. In this case, the signal
from Pair 5 was subtracted from that of Pair 3, resulting in
increased signal amplitude. Reprinted by permission from
Reference 19.
550
est to the crural diaphragm at any point in the respiratory cycle. Application of a double subtraction technique using the
difference between signals obtained from each electrode pair
caudal and cranial to the crural diaphragm (19) further enhances the signal-to-noise ratio.
Advantages of esophageal recordings are that they are not
influenced by obesity and that, when used with or combined
into the same catheter as esophageal and gastric pressure monitors (12), they enable simultaneous recordings of diaphragm
EMG and transdiaphragmatic pressure. Disadvantages include
the discomfort and the remote risks of regurgitation, aspiration, and vagally mediated bradycardia associated with their
placement. Diaphragmatic hernia may be a source of error in
esophageal recordings. In theory, the EMG from esophageal
electrodes may not be representative of the diaphragm as a
whole, because it samples only the crural portions of the diaphragm. However, although cruralcostal dissociation during
breathing has been demonstrated in animals (20, 21), it appears not to be a problem in humans (2224).
Intramuscular electrodes: advantages. Intramuscular electrodes provide relatively selective recordings from nondiaphragm respiratory muscles (911, 2530) with sufficient discrimination of individual motor unit activity to allow evidence
of denervation or myopathy to be detected. Whitelaw and
Feroah (30) have provided a detailed description of a safe
method to record single motor unit activity in intercostal muscles with wire electrodes. A number of investigators have
demonstrated techniques for placement of monopolar or bipolar needle electrodes in the human diaphragm, either by a
medial subcostal approach (31, 32) or by a lower intercostal
approach (33, 34) (Figure 2), sometimes assisted by real-time
ultrasound. Fine wire electrodes for single motor unit recordings have also been implanted in the right hemidiaphragms of
humans (13, 35). Although the flexible nature of wire electrodes makes them relatively stable during volume changes of
up to 1.5 L around the FRC, artifactual changes in recording
conditions occur with larger volume changes (13). Wire im-
Figure 2. Intercostal approach to placement of an electromyography

needle in the diaphragm. The needle is inserted one interspace above
the lower costal margin between the anterior axillary and medial clavicular lines. During quiet respiration, the needle passes well away
from the visceral pleura. Reprinted by permission from Reference 34.
VOL 166
2002
plants are often used in upper airway muscle studies (see

ELECTROMYOGRAPHY in Section 8 of this Statement).
As far as advantages and disadvantages are concerned, intramuscular electrodes are optimal for analysis of action potentials in the assessment of myopathic changes and for comparing single motor unit firing frequencies among different
respiratory muscles and clinical conditions (13) (Figures 3 and
4). Cross-talk is less of a problem with implanted than surface
electrodes, although it is not completely eliminated. Implanted electrodes are difficult to place, however, and are potentially less useful than surface or esophageal recordings for
quantifying global respiratory muscle activity. For intercostal
recordings and especially for triangularis sterni and diaphragm recordings (26), there is a small risk of pneumothorax.
As with any skin-penetrating technique, there are risks of
bleeding and bruising. With the use of sterile disposable needles or wires, there is no practical risk of transmission of infectious disease.
EMG signal processing. The preferred amplifier design for
detection of weak myoelectric signals is a differential amplifier, which amplifies the difference between two paired inputs
and thereby eliminates signals, such as 50 or 60 Hz from power
lines, that have common influences on both outputs (common
mode rejection). Input impedance is high to minimize loss of
voltage across the two active electrodes. Bandpass filtering is
usually employed, with a high-pass filter to remove the signals
direct current component and a low-pass filter set below the
data acquisition frequency to avoid distortion (by aliasing) of
the signal caused by undersampling. Filter settings will vary,
depending on the type of electrode and the application; for
surface or esophageal recordings, a band width of 10 to 1,000
Hz is often used.
For simple measurements of motor latencies or CMAP amplitudes, the data display can be as simple as a storage oscillo-
Figure 3. (A) A needle EMG record from a normal diaphragm at 200

milliseconds/division and 200 mV/division (B) The same subject recorded at 10 milliseconds/division and 200 mV/division. Reprinted by
permission from Reference 34.
551
Figure 4. (A) Needle EMG of the diaphragm of a patient with porphyric neuropathy, showing a few fibrillation potentials and positive wave
potentials, along with QRS complexes (calibration, 10 milliseconds/division and 100 mV/division). (B) Needle EMG of the same patient 4
weeks later, showing increased numbers of positive sharp waves and
fibrillation. Reprinted by permission from Reference 34.
scope equipped with a camera. A more versatile system, essential for frequency domain measurements, is a computer with
an analog-to-digital converter. Because in digitally sampled
signals only spectral components of frequencies lower than
half the sampling frequency can be observed (the Nyqvist theorem), the sampling frequency should be chosen accordingly.
To provide an analog signal proportional to average
EMG activity at any point in time, many investigators subject
the raw EMG signal to rectification and leaky integration.
Unless sophisticated analog gating and/or filtering techniques
are employed, however, the effects of multiple artifacts detailed in the following sections limit this technique to providing only a crude estimation of the level of muscle activation.
Data Analysis
Time domain EMG analysis. Time domain EMG analysis represents the electrical activity of the muscles as a function of time. At
low levels of contraction, isolated MUAPs can be distinguished
and analyzed by such time domain indices as signal amplitude and
different types of integrated EMG, including full wave rectified
and averaged signal (FRA) and root mean square of the signal
(RMS) (36), all of which increase as a function of the number of

motor fibers in the unit. Frequency-related indices, such as rise
time and duration, can also be measured (36, 37); these are influenced more by action potential conduction velocity and the width
of the innervation zone than by the size of the motor unit.
With increasing levels of contraction, more motor units are
recruited, and the firing rate increases. The resulting interference pattern EMG can be analyzed by indices such as amplitude, RMS, and FRA, and by frequency-related indices such
as zero crossing distance and turn-point distance (36).
Frequency domain analysis. Frequency domain analysis is a
technique to express EMG power as a function of frequency.
EMG power is intimately related to its frequency characteristics
(8), and frequency characteristics in turn are related to muscle
membrane conduction velocity, to filtering properties of the
electrodes, to muscleelectrode distances, and to noise (36). Frequency domain analysis greatly simplifies the evaluation of all of
these factors, which are difficult to evaluate in the time domain.
The relative contributions of high and low frequencies to
EMG signals can be estimated crudely by splitting the signal,
filtering the signal with different bandpass filters, integrating
the output of each bandpass filter, and comparing the two integrals, to yield a ratio of high-frequency power (e.g., 130250 Hz)
to low-frequency power (e.g., 3050 Hz), the H/L ratio (38).
More useful information can be obtained by power spectral
analysis, for which a window of time domain EMG data is
digitized and subjected to a computerized fast Fourier transform. The fast Fourier transform components are squared and
their products are calculated, giving the power spectrum, which
graphs the power of the signal as a function of frequency (Figure 5). To avoid artifactual overestimation of high-frequency
content, it is necessary to condition the window of data by tapering its amplitude at the beginning and end of the window or
by replacing all data before the first and after the last zero
crossing with zeros. These shaping processes lead to a slight
underestimation of high-frequency content of the EMG (39).
Numerical quantification of the power spectrum is possible
by calculating the spectral moments (36). Spectral moments
(M) of order n are defined as:
i max
Mn =
P f
i
n
i
i=0
where P is power density, f is frequency, i is the index over

which the power density product is summed, i 0 is the direct
current component, and imax is the index associated with the
highest frequency in the spectrum.
The spectral moment of zero order (M0) and the root mean
square (RMS M 1/2
0 /p, where p is the number of points in the
sample) are indices of total EMG power. Theoretically, RMS
reflects the force output of the muscle (8). However, both M0
and RMS are influenced by a number of other parameters, especially conduction velocity (8). In applications in which the
power spectrum is expected to shift, as in fatigue, the first-order
spectral moment (M1) may be a more useful index of muscle
activation, because it is not affected by changes in action potential conduction velocity (36). Quantification of the distribution of power in the spectrum can be obtained by calculating
the center frequency (fc M1/M0), also known as the mean or
centroid frequency. Figure 5 shows how EMG signal power
spectra are influenced by interelectrode distance.
Artifacts
Filter effects. Muscle-to-electrode distance. Increasing muscleto-electrode distance results in reduced signal amplitude with
relatively larger attenuation of high- than low-frequency power,
552
VOL 166
2002
Figure 5. Time and frequency domain characteristics of diaphragm EMG signals obtained with bipolar esophageal electrodes
positioned near a region with low density
of motor end-plates and aligned in the direction of diaphragm fibers. Left: Raw EMG
and computed power spectrum from a
pair of electrodes with a 10-mm interelectrode distance. Right (top and bottom):
Signals derived at the same time from the
same region using electrodes with a 5-mm
interelectrode distance. Right (middle): On
the left is shown a cross-correlogram obtained from the successive cross-correlations at various time delays between EMG
signals obtained with a 5-mm interelectrode distance; on the right is shown two
motor unit action potentials (MUAPs) obtained from the two electrodes pairs with
a 5-mm interelectrode distance. Reprinted
so that the relative distribution of power in the spectrum is altered and fc is reduced (1, 1417, 40) (Figure 6).
Interelectrode distance and alignment with respect to fiber
direction (bipolar arrangement). Both interelectrode distance
and the orientation of the electrodes with respect to fiber direction can affect power spectra (1, 18, 41). As shown in Figure 5, reductions in interelectrode distance reduce signal
Figure 6. The influence of esophageal electrode positioning on diaphragm EMG center frequency (fc) (left axis; circles) and root mean
square (RMS) (right axis; squares). Solid symbols indicate experimental
data, whereas open symbols indicate computer-simulated power spectra. The 0 on the x axis indicates the center of the electrically active
region (source). RMS was lowest at the electrode pair positioned directly over the source (due to the cancellation effect of a bipolar electrode). Caudal and cephalad to the source, RMS increased progressively to maxima at approximately 10 mm from the source in either
direction. fc was maximal at the electrode pair directly over the source,
hence less high frequency is filtered by the electrode being closer to
the muscle. Reprinted by permission from Reference 46.
power, with relatively higher attenuation of low-frequency

components. With bipolar recordings of EMG, electrode orientation with respect to fiber direction and interelectrode distance should be standardized, and investigators should recognize that these relationships are likely to change with muscle
contraction.
Signal disturbances. Electrode motion-induced artifacts.
Movement of the electrode or a change in the pressure on the
electrode results in a redistribution of the charges in the resistivecapacitive interface between the electrode and the tissues, causing relatively large amplitude artifacts with low frequency (mostly below 2025 Hz) (39, 42). Most of the motion
artifact can be filtered out with high-pass filters. However,
some EMG power occurs at frequencies below 25 Hz, and
even the most efficient filter attenuates some of the power
above its cutoff frequency. Thus, high-pass filtration inevitably leads to loss of low-frequency power from the EMG signal.
When power spectrum analysis is applied, the power related
to electrode movement can be reduced (42) or replaced by an
extrapolation of the diaphragm EMG power to those frequencies (39).
Noise. Background noise, signals of unidentifiable origin,
can be assumed to have constant power density over the frequency region of interest in EMG recordings. On the basis of
this assumption, the noise component of the signal-to-noise
ratio is usually estimated from power density values obtained
in the uppermost frequency range of the EMG power spectrum (39, 43, 44). Noise originating from modern electrophysiologic instruments is not usually a problem, but ancillary
equipment, such as pressure or motion sensors, pumps, or respirators, may introduce noise in various frequency ranges.
Disturbances from power lines. Sinusoidal alternating currents (ACs) originating from power lines can usually be reduced to a negligible level by proper shielding of electrode cables and connections, by using amplifiers with high common
553
mode rejection ratio, and by connecting all instruments to the

same ground point. Residual power line disturbances can be
filtered out with notch filters, or, alternatively, the power of
the affected frequency and its harmonics can be excluded in
power spectral analysis.
Potential artifactual influences of physiologic origin. Crosstalk signals. These are signals originating from muscles other
than the muscle being investigated. The best-described source
of cross-talk is electrical activity from the heart (EKG).
Esophageal recordings of diaphragm EMG are particularly
susceptible to cardiac cross-talk, which provides a signal with
some 10 times the power of the diaphragm EMG but with a
much lower fc (39, 42). For measurements of stimulation-elicited CMAPs, EKG artifacts are relatively easy to avoid by
triggering the stimulus from the QRS complex with an appropriate delay. Alternatively, any record visibly superimposed
on the QRS complex can simply be discarded.
For time domain analyses over long periods of time or for
power spectral analyses, EKG artifacts pose a significant
problem. Heart rates often change, as, for example, with inspiration and with exercise. The frequency content of the EKG is
lower than that of EMG activity, but there is considerable
overlap (39, 43). Thus, if the high-pass filter is set high enough
to eliminate most of the EKG activity, it will also eliminate
much of the EMG activity. When EMG frequency content
might be shifting toward lower frequencies, as during increasing muscle effort, excessive high-pass filtration will result in
spurious reductions in measured RMS. Methods to eliminate
the cardiac activity from esophageal recordings of diaphragm
EMG by template subtraction (45) have been proposed. However, because the amplitude and shape of the electrocardiogram change with changes in lung volume, the safest method
to avoid cardiac influences on diaphragm EMG signals still
appears to be selection of signal segments that are free of cardiac activity (39).
Another source of cross-talk in esophageal recordings of
diaphragm EMG is esophageal peristalsis (39). The esophageal peristalsis signals show a relatively large amplitude and
relatively low-frequency content. In general, signal segments
contaminated by esophageal peristalsis, usually easily identified as strong, slow esophageal pressure waves, should be excluded from analysis.
Other examples of cross-talk sources include contamination of diaphragm EMG data recorded from electrodes on the
lower rib cage with abdominal or intercostal muscle activity
and contamination of intercostal muscle EMGs by interference from pectorals, abdominals, or diaphragm (46).
Innervation zones. Electrodes positioned close to or over
an innervation zone (a region with a high density of motor end
plates) produce complex interference patterns, because the

action potentials elicited by firing of an individual motor unit
may propagate in opposite directions relative to the electrodes
(2, 47, 48). Characteristically, there is reduced total power and
increased high-frequency content (49). This effect is maximal
when bipolar electrodes are oriented parallel to fiber direction; esophageal electrodes, which are arranged approximately perpendicular to fibers, are less susceptible (18).
Influence of changes in muscle length or chest wall configuration. Changes in chest wall configuration systematically affect the amplitude and frequency content of evoked CMAPs
measured with chest surface or esophageal electrodes (13, 50).
With spontaneous EMG activity, artifact-free signals recorded
from esophageal electrodes are not systematically affected by
changes in chest wall configuration (17, 18, 25, 50, 51). Because conduction velocity is not significantly affected by muscle length (52), most of the chest wall configuration effects on
EMG are probably due to muscle-to-electrode distance or orientation changes.
Influence of changes in muscle temperature. Muscle temperature increases during exercise, because of the increase in
blood flow to the muscle (53, 54), and metabolic heat production (55). Because the propagation velocity of the muscle fiber
action potential is correlated with temperature (56, 57), EMG
frequency content must also be temperature dependent. No
method for control of or correction for temperature in recordings of respiratory muscles has yet been described.
Applications
See Table 2.
Single fiber and motor unit analysis. Single fiber and motor
unit signal analyses are useful for diagnosis of nerve or muscle
pathology. For the diagnosis of neuromuscular disease, limb
muscles are more commonly studied than respiratory muscles,
because they are more readily accessible. A number of investigators have demonstrated the usefulness of needle electromyography of the diaphragm for the diagnosis of neuromuscular
disease, particularly neuropathic processes such as Guillain
Barr syndrome, lower motor neuron involvement with spinal
cord injury, and polyneuropathy of critical illness (31, 32, 34).
Bolton (34) has pointed out that the relatively high-frequency, low-amplitude potentials of the normal diaphragm
are often difficult to differentiate from myopathic potentials.
Nevertheless, several neuromuscular diseases present primarily with respiratory muscle weakness; as experience is gained
with single fiber and motor unit analysis of respiratory muscles, these techniques applied to respiratory muscles may provide the earliest evidence of the neuro- or myopathic process.
TABLE 2. APPLICATIONS FOR RESPIRATORY MUSCLE ELECTROMYOGRAMS

Type of Test
Needle EMG
Interference pattern EMG signal
Condition
Denervation
Demyelination
Chronic denervation
Myotonia
Myopathy
Paralysis or dyscoordination
Quantification of neural drive
Efficiency of contraction
Fatigue
Finding
MUAPs, fibrillation potentials, and
positive sharp waves
MUAPs, without potentials
No., size of MUAPs
Myotonic discharges
Short, polyphasic potentials
Respiratory muscle activation pattern
Changes in FRA or RMS
Pdi/Edi
Spectral analysis
Definition of abbreviations: Pdi/Edi ratio of tidal respiratory change in transdiaphragmatic pressure to tidal respiratory change in integrated diaphragm EMG; EMG electromyography; FRA full wave rectified and averaged signal; MUAP motor unit action potential;
RMS root mean square of the signal.
554
Interference pattern signal. The interference pattern EMG

(raw EMG from surface electrodes) of respiratory muscles is
useful for the determination of the timing and level of muscle
activation during respiratory activities. Thus, EMGs can help to
determine which of the many respiratory muscles are active in
various phases of respiration, in various body positions, in various states of consciousness, and in various clinical conditions
(see ELECTROMYOGRAPHY in Section 8 of this Statement). Specifically, the absence of voluntary or involuntary EMG activity
can be used as evidence of paralysis of specific respiratory muscles. EMGs can also help to quantify the respiratory muscle activation responses to loaded breathing and to CO2-stimulated
breathing or to monitor and control mechanical ventilation (58,
59). Furthermore, when related to pressure or force developed
by respiratory muscles, EMGs can help to assess the electromechanical efficiency of respiratory muscle function (see ELECTROMECHANICAL EFFECTIVENESS in Section 6 of this Statement).
Interindividual comparisons of absolute FRA or RMS values do not meaningfully reflect respiratory drive, because of
varying filtering influences of electrode placement relative to
the contracting muscles and/or anatomic differences between
subjects, for example, the amount and type of interlaying tissue. However, changes in these indices in response to interventions such as changes in inspired CO2 concentration, loaded
breathing, exercise, states of consciousness, drugs, or other influences, reflect changing motor output of the central nervous
system (CNS) to respiratory muscles. Because some respiratory muscles are silent during quiet breathing, it is not practical to normalize respiratory muscle EMG activity to that observed during resting tidal breathing. It is often more practical
to normalize EMG activity to that observed during sniff inhalations or maximal inspiratory efforts (22).
The interference pattern EMG of respiratory muscles may
also be useful for the assessment of respiratory muscle fatigue
(see ELECTROMYOGRAPHY in Section 5 of this Statement). Localized muscle fatigue is accompanied by a reduction in muscle fiber action potential conduction velocity (1), which is reflected
in the diaphragm EMG power spectrum as a shift toward lower
frequencies (41, 52). This spectral shift is most commonly
quantified as a reduction in fc. Spectral shifts have been detected in healthy subjects during inspiratory resistive breathing
(38, 60), in patients with weak inspiratory muscles constrained
to breathe with prolonged inspiratory duty cycles (61) or with
exertion-induced inspiratory muscle overload (62), and in patients with chronic obstructive pulmonary disease during exertion (43), and they are associated with changes in respiratory effort sensation (63). The EMG frequency spectrum is often
influenced by other factors, such as the signal-to-noise ratio,
electrode position, and recruitment of muscles that potentially
STIMULATION TESTS
Rationale
Peripheral nerve, spinal, or cortical stimulation, either by implanted electrodes (for peripheral nerves) or by externally applied electric or magnetic fields, elicit relatively synchronized
activation of motor units at reproducible and predictable levels. The resulting compound action potentials and subsequent
muscle contraction allow for measurement of the efficiency of
neural and neuromuscular transmission. The muscle responses
to stimulation are discussed in PHRENIC NERVE STIMULATION in
Section 2 of this Statement.
Scientific Basis
For practical purposes, human respiratory motor nerves are not

accessible over a sufficient length to permit phrenic or other respiratory nerve conduction velocities to be measured. However,
motor latency can be measured. Included in the latency are
the times required for (1) initiation of action potentials in the
axons, (2) rapid saltatory conduction through myelinated axons, (3) slow conduction along the thinner terminal twigs, and
(4) chemical transmission across the neuromuscular junctions.
Equipment
Nerve stimulation. Stimulation of respiratory nerves can be accomplished with electrical or magnetic stimulators (Table 3).
The former are less expensive, less cumbersome, more rugged,
and more precisely controllable; the latter are easier to apply and
less painful for patients. For electrical stimulation, the phrenic
nerve is stimulated transcutaneously by surface electrodes at the
posterior border of the sternomastoid, or with implanted needle,
wire, or hook electrodes. The phrenic nerve(s) can also be stimulated magnetically, via a dorsal cervical approach, which stimulates the lower cervical nerve roots; via an anterior presternal approach, which stimulates both phrenic nerves; or, using one or
two figure-of-eight coils, unilaterally or bilaterally over the anterior neck to stimulate one or both phrenic nerves.
Several other respiratory nerves and muscles can also be
stimulated, either transcutaneously or by needle or wire electrodes. Pradhan and Taly (28) have demonstrated a technique
for stimulating lower intercostal nerves via probe electrodes
for latency measurements. The ventral roots of intercostal
nerves have been stimulated either by high-voltage stimulation over the spine (6466) or by surgically implanted wire elec-
Advantages
Disadvantages
Electrical
Implanted
Less uncomfortable during stimulation

Precise control of stimuli
Difficult to place
Transcutaneous
Inexpensive
Requires little preparation
Painful
Difficult to maintain contact
during voluntary effort
Does not require contact with skin

Painless
Expensive and cumbersome

Relatively unselective
Difficult to stimulate repetitively
Precise location of stimulus uncertain
(so latency measurements may be
problematic)
Magnetic
2002
contribute to cross-talk. Technical factors affecting the EMG

power spectrum must always be taken into account before any
physiologic interpretations of a spectral shift can be made (1,
1519, 4148).
TABLE 3. TYPES OF RESPIRATORY MUSCLE STIMULATION

Type of Stimulation
VOL 166
trodes (67). The rectus abdominus and oblique muscles can

also be stimulated with large surface area electrodes (68), and
abdominal muscle stimulation has been shown to be effective
enough to facilitate cough in tetraplegic patients with impaired expiratory muscle function (66). The abdominal muscles can be activated by magnetic stimulation of nerve roots at
the level of T10.
In most cases, it is important to be sure that the delivered
stimulus is strong enough to activate maximally all motor units
in the muscle of interest. To that end, elicited CMAP amplitude is measured as a function of stimulus intensity, and, for
subsequent measurements, stimulus intensity is set to be supramaximal, 2050% above that required for maximal response. During long experiments, it should be regularly verified that stimulus intensity is supramaximal.
Cortical stimulation. It is now possible to stimulate cortical
and subcortical neural pathways in human subjects, using highvoltage (up to 1,500 V) electrical stimulators (69) and magnetic
stimulators (up to 3.0 T) (70). Electrical stimulation of the cortex requires saline-soaked gauze pads (or silversilver chloride
electrodes) on the scalp with the anode positioned over the relevant region of the motor cortex. For the diaphragm and intercostal muscles, the optimal site is close to the midline at or just
anterior to the vertex (71). The cathode can either be a single
electrode 68 cm anterior to the anode, a ring of electrodes
placed around the scalp, or an electrode several centimeters lateral to the anode at the vertex. Commercially available magnetic stimulation coils consist of several circular coils in a single
housing (usually 10 cm in diameter) positioned (usually tangentially) with their edges close to the scalp region of interest.
Positioning of large circular coils slightly anterior or posterior
to the vertex will routinely activate corticofugal output from the
medial parts of both hemispheres, including major inspiratory
and expiratory muscles in normal subjects, and structures as lateral as the hand area are also activated. More focal stimulation
can be achieved with double or butterfly coils. There are several nonstandard coils, but their properties must be carefully
determined before their clinical utility can be assessed.
To obtain minimal latencies for response to transcranial stimulation, and thereby to avoid coming to the erroneous conclusion
that there is a deficit in corticospinal conduction, it is best to
record them during voluntary (or reflex) contractions producing
at least 10% of maximal force. Responses can be obtained without a background contraction, but the latencies are more variable.
The presence of cardiac or other implanted electrical devices, including pacing wires and pulmonary artery catheters
equipped with thermisters, is an absolute contraindication for
magnetic stimulation and for high-voltage electrical stimulation of cranial or spinal structures. Epilepsy and known intracortical pathology are relative contraindications for cranial
stimulation, as is the presence of intracranial clips.
Although transcranial stimulation has now been used for
well over a decade with few reports of side effects, two issues
deserve mention. First, there is a risk of induction of a seizure,
particularly in those with pre-existing cortical pathology (such
as a recent cerebrovascular accident). Although seizures have
not been observed with repeated single stimuli, greater precautions are necessary for the newer rapid-rate stimulators
(72, 73). Second, the use of some but not all transcranial magnetic stimulators has caused sustained elevations in auditory
threshold, presumably due to the brief, but high-intensity
click produced by the stimulus passing into the coil.
Data Analysis
Nerve stimulation is essential for measurements of nerve conduction velocity or latency. Furthermore, the electromyographic
555
signal elicited by supramaximal nerve stimulation (CMAP)

provides a different perspective than spontaneous EMG, with
two distinct advantages: assurance of maximal activation, and
a generally higher signal-to-noise ratio. CMAPs are detected,
usually with surface electrodes applied over the costal margin,
and the time between triggering of the stimulus and detection
of the elicited CMAP is recorded. In most cases, CMAP amplitude and/or area are also recorded.
Latencies can also be measured after cortical stimulation
(Figure 7). The central conduction time (CCT) is an indirect
estimate of the time taken for the descending volley to travel
from the motor cortex to the relevant motoneurons. The CCT
can be measured by subtraction of the peripheral conduction
time (estimated by stimulation over the relevant spinal segment or root) from the total conduction time, from stimulus to
onset of the motor-evoked potential (MEP). The measured
CCT, however, is not a true measurement of central conduction velocities, because the exact site and timing of the relevant descending corticofugal volleys show some variability
and because stimulation over the spinal cord activates the motor axons at a variable distance from the motoneurons. The
MEP observed in EMG recordings of most human trunk and
neck muscles, including diaphragm, intercostals, scalene, and
abdominal muscles (64), has an onset latency consistent with a
rapidly conducting oligosynaptic pathway. There is no evidence that this MEP involves a contribution from bulbopontine respiratory neurons.
The MEP can also be assessed in terms of amplitude (and
dispersion). Transcranial electrical stimulation via an anode
over the appropriate scalp site evokes a direct corticospinal
volley (D-wave) followed by a series of indirect trans-synaptically evoked corticospinal volleys (I-waves; interval between
volleys, 0.81.0 milliseconds). With increased intensity of electrical stimulation, the site of activation along the corticospinal
path is located deeper within the brain, reaching the level of
the pyramidal decussation in the medulla with strong stimuli
(74). The size of the D-wave increases with increased stimulus
intensity, as does the size of I-waves. With transcranial magnetic stimulation, the precise corticospinal response depends
on the location of the evoked currents. Hence, there are differences in the magnetic activation of the corticospinal output
from the hand areas (lateral region of the primary motor cortical strip) compared with the leg areas (within the medial edge
of the motor cortex) with different directions of stimulus current. With high-power nonfocal coils (which have the greatest
diagnostic utility) transcranial magnetic stimuli evoke not only
Figure 7. Comparison of motor latencies and compound diaphragm

or deltoid action potentials elicited by magnetic stimulation at the
phrenic nerve, spinal level, and cortical level. The latencies are indicated in milliseconds. Bil. bilateral; C4 level of the 4th cervical
root. Reprinted by permission from Reference 64.
556
VOL 166
2002
Applications
Figure 8. Effects of cortical stimulation at rest and during voluntary inspiratory efforts of varying strength. Left: Transdiaphragmatic pressure
(Pdi) twitches obtained by bilateral supramaximal phrenic nerve stimulation during relaxation (thick trace). The thin traces show twitches resulting from transcranial magnetic stimulation (stimulator set at 85%
maximal) at rest (lowest trace, no twitch) and during three graded
static inspiratory efforts. Note the facilitation of Pdi twitches at intermediate voluntary efforts and the partial occlusion at high effort. Right:
Compound muscle action potentials (CMAPs) recorded from surface
electrodes on right and left chest wall (over the diaphragm) during
magnetic stimulation at the four levels of voluntary effort shown at
left. Note the progressive facilitation in CMAP amplitude. Reprinted by
permission from Reference 76.
I-waves but also small D-waves (75). There is probably more

trial-to-trial variability in the components of the descending
volley to magnetic stimulation than to electrical stimulation.
During voluntary contraction, the minimal latency of responses is reduced, and the MEPs are increased in size compared with relaxation (Figure 8), reflecting increased excitability not only at the spinal level (76) but also at the motor cortex
(77). For the diaphragm, this effect has been documented during both volitional efforts (35) and CO2-driven hyperpnea (78).
Interpretation of the results of motor cortical stimulation must
be made with caution for several reasons. First, stimulation over
the motor cortex activates both excitatory and inhibitory structures within the cortex. Second, a single stimulus as brief as 50
milliseconds evokes multiple descending corticospinal volleys.
Third, the evoked motor response is elicited preferentially in
motoneurons that are close to firing threshold as a net result of
voluntary, involuntary, and reflex inputs. Finally, change in MEP
can be said to reflect a change in cortical physiology only if the excitability of the spinal cord has been proven to remain constant.
When cortical stimulation is delivered during voluntary effort, the MEP is followed by a period of near silence in the
EMG recorded from the relevant muscle. The latter part of
the silent period is due to inhibition of motor cortical output
(7981). The duration of the silent period grows with increasing intensity of cortical stimulation but does not vary greatly
with the relative strength of voluntary contraction.
Normal phrenic nerve/diaphragm latencies, elicited by electrical stimulation at the neck, have been reported to average 68
milliseconds in adults (12, 16, 82, 83), with lower values in children (see Table 4). Because the right phrenic nerve is shorter
than the left, latency is slightly shorter on the right side.
CMAP amplitudes, recorded from chest wall surface electrodes, average 500800 mV. Normal values for magnetic
stimulation-induced phrenic nerve/diaphragm latency or for
intercostal nerve/muscle latency have not been fully established (84, 85).
Phrenic nerve/diaphragm latencies are abnormally slow in
demyelinating polyneuropathies, notably Guillain-Barr syndrome. They are usually nearly normal, but are associated
with markedly depressed CMAP amplitude, in traumatic neuropathies, such as postcardiac surgery phrenic nerve palsy or
the polyneuropathy of critical illness. In myasthenia gravis, a
reversible decrement in diaphragm CMAP can be elicited by
repetitive phrenic nerve stimulation.
CCTs from cortex to phrenic motoneurons are approximately 4 milliseconds in normal adults (35). Normal values for
D- and I-wave amplitudes and the effects of disease are not
yet established, but it is apparent that I-wave amplitude is reduced by general anesthetic agents.
Transcranial stimulation to determine CCT has been used
in the assessment of a range of upper motor neuron disorders
and particularly in the assessment of patients with possible
CNS demyelination, including multiple sclerosis. It can be applied specifically to the respiratory muscles, or, more commonly, to muscles in both upper and lower limbs.
Compound muscle action potentials (CMAP) in response
to electrical or magnetic nerve or cortical stimulation can also
provide useful information. Lack of a CMAP after nerve stimulation is an indication of paralysis, with the lesion located
proximal to or at the neuromuscular junction. Lack of a
CMAP in response to cortical stimulation when a CMAP is
elicited by phrenic nerve stimulation has been used to identify
good candidates for phrenic nerve pacing.
Changes in CMAP amplitude, especially as compared with
changes in elicited muscle twitch strength (such as phrenic
stimulation-induced diaphragm CMAP as compared with
transdiaphragmatic twitch pressure [Pdi,tw]) can be used as
evidence for or against the development of neural or neuromuscular transmission defects (when both Pdi,tw and CMAP
decrease) or contractile defects (when Pdi,tw decreases but
CMAP does not) (86).
CONCLUSION
In a manner analogous to the use of the electrocardiogram to
assess cardiac function, electrophysiological tests of respiratory musclesrespiratory muscle motor latencies and elec-
TABLE 4. APPLICATIONS OF RESPIRATORY MUSCLE STIMULATION TESTS

Stimulation Test
Normal Values (in Adults)
Abnormal in:
Phrenic nerve/diaphragm latency
68 ms
Demyelinating polyneuropathies, e.g.,

GuillainBarr syndrome
Elicited CMAP
Depends on recording electrode
Phrenic nerve palsy, neuromuscular

transmission defect, polyneuropathy
of sepsis, myopathies
Central conduction time, total

conduction time minus
phrenic nerve latency
4 ms
Multiple sclerosis
Definition of abbreviation: CMAP compound muscle action potential.
557
tromyographycan be used to assess (1) respiratory drive, respiratory muscle coordination, and the level of activation of
individual muscles; (2) the presence of neural and neuromuscular pathology; and (3) the apparent efficacy of the contractile function of the muscles, when used in conjunction with
measurements of contractile force. The special challenges presented by data analysis complexity and by a host of potential
artifacts lead to the need for great care in the application of
EMG techniques to respiratory muscles. Nevertheless, neurophysiological tests are emerging as indispensable components
of the respiratory muscle physiologists arsenal.
SUMMARY
This Section of the Statement has described available electrophysiologic tests, the functions of which are to assess the integrity of the respiratory neuromotor apparatus. These electrophysiologic tests are technically complex and require considerable
expertise.
There are two main types of test: electromyography (EMG)
and stimulation tests.
Type 1: EMG. For the respiratory muscles the EMG can
be used to assess the level and pattern of activation, to detect
and diagnose neuromuscular pathology, and when combined
with tests of mechanical function to assess the efficacy of contraction.
The EMG can be recorded with surface electrodes (for diaphragm, intercostal, scalene, abdominal, and accessory muscles) or an esophageal electrode (for the crural diaphragm).
Surface electrodes are noninvasive and sample a large number
of motor units, but contamination (cross-talk) from other
muscles is a substantial problem, as is the effect of body size
and shape on signal amplitude. Esophageal electrodes provide
more specific information, but the technique is invasive and
complex.
Surface and esophageal electrodes can record the interference pattern EMG (raw EMG) of the respiratory muscles and
are useful to determine the timing and level of respiratory
muscle activation during breathing, at rest, and under load.
Frequency domain analysis of the EMG is used, as a research
tool, to investigate respiratory muscle fatigue (discussed in
Intramuscular electrodes can be used to record, relatively
selectively, from the diaphragm and intercostal muscles. Motor neuron firing frequency can be measured and neuromuscular disorders diagnosed. However, the techniques are invasive and technically difficult.
Type 2: Stimulation Tests. Stimulation tests measure the
efficiency of neural and neuromuscular transmission.
Nerve stimulation can be achieved with electrical or magnetic stimulators. Electrical stimulation is inexpensive and relatively selective, but is uncomfortable and can be technically
difficult. Magnetic stimulation is easier to achieve and less uncomfortable, but can be less selective and is expensive.
Most commonly the phrenic nerves are stimulated and the
diaphragm EMG elicited, for the measurement of phrenic
nerve/diaphragm latencies and CMAP amplitudes. Latencies
are prolonged in some neuromuscular disorders (e.g., demyelination) and CMAP is reduced in amplitude (e.g., traumatic
damage to the phrenic nerves).
Cortical stimulation is most commonly performed with a
magnetic stimulator, and permits the measurement of central
conduction times (CCT) for limb muscles and diaphragm.
CCT can be prolonged as, for example, in multiple sclerosis.
Cortical stimulation is not selective, and the application of the
technique to the respiratory system is a highly specialized skill.
References
1. Lindstrm L, Magnusson R. Interpretation of myoelectric power spectra: a model and its applications. Proc IEEE 1977;65:653662.
2. Lindstrm L. A model describing the power spectrum of myoelectric signals.
I. Single fiber signal. Res Lab Med Electr, Gteborg, Sweden. 1973;5:73.
3. Stlberg, E. Propagation velocity in human muscle fibers in situ. Acta
Physiol Scand 1966;70(Suppl 287):2112.
4. Hkansson CH. Conduction velocity and amplitude of the action potential as related to circumference in the isolated frog muscle. Acta Physiol Scand 1956;37:1434.
5. Juel C. Muscle action potential propagation velocity changes during activity. Muscle Nerve 1988;11:714719.
6. Metzger JM, Fitts RH. Fatigue from high- and low-frequency muscle stimulation: role of sarcolemma action potentials. Exp Neurol 1986;93:320333.
7. Broman H, Lindstrm L. A model describing the power spectrum of
myoelectric signals. II. Motor unit signals. Res Lab Med Electr, Gteborg, Sweden. 1974;8:74.
8. Lindstrm L, Broman H. A model describing the power spectrum of
myoelectric signals. III. Summation of motor unit signals. Res Lab
Med Electr, Gteborg, Sweden. 1974;9:74.
9. De Troyer A, Peche R, Yernault JC, Estenne M. Neck muscle activity in
patients with severe chronic obstructive pulmonary disease. Am J
10. De Troyer A, Estenne M, Ninane V, Van Gansbeke D, Gorini M. Transversus abdominis muscle function in humans. J Appl Physiol 1990;68:
10101016.
11. De Troyer A, Farkas GA. Linkage between parasternals and external
intercostals during resting breathing. J Appl Physiol 1990;69:509516.
12. McKenzie DK, Gandevia SC. Phrenic nerve conduction times and twitch
13. Gandevia SC, McKenzie DK. Human diaphragmatic EMG: changes
with lung volume and posture during supramaximal phrenic nerve
stimulation. J Appl Physiol 1986;60:14201428.
14. Grassino AE, Whitelaw WA, Milic-Emili J. Influence of lung volume
and electrode position on electromyography of the diaphragm. J Appl
Physiol 1976;40:971975.
15. Kim MJ, Druz WS, Danon J, Machnach W, Sharp JT. Effects of lung
volume and electrode position on esophageal diaphragmatic EMG. J
Appl Physiol 1978;45:392398.
16. nal E, Lopata M, Ginzburg AS, OConnor TD. Diaphragmatic EMG
and transdiaphragmatic pressure measurements with a single catheter.
17. Beck J, Sinderby C, Weinberg J, Grassino AE. Effects of muscle-to-electrode distance on the human diaphragm electromyogram. J Appl Physiol 1995;79:975985.
18. Beck J, Sinderby C, Lindstrm L, Grassino A. Influence of bipolar electrode positioning on measurements of human crural diaphragm EMG.
J Appl Physiol 1996;81:14341439.
19. Sinderby CA, Beck JC, Lindstrm LH, Grassino AE. Enhancement of
signal quality in esophageal recordings of diaphragm EMG by the
double subtraction technique. J Appl Physiol 1997;82:520530.
20. Van Lunteran E, Haxhiu MA, Cherniak NS, Goldman M. Differential
costal and crural diaphragm compensation for posture changes. J
Appl Physiol 1985;58:18951900.
21. Sprung J, Deschamps C, Hubmayer RD, Walters B, Rodarte J. In vivo
regional diaphragm function in dogs. J Appl Physiol 1989;67:655662.
22. Sinderby C, Beck J, Weinberg J, Spahija J, Grassino A. Voluntary activation of the human diaphragm in health and disease. J Appl Physiol
1998;85:21462158.
23. Beck J, Sinderby C, Lindstrom L, Grassino A. Effects of lung volume on
diaphragm EMG signal strength during voluntary contractions. J Appl
Physiol 1998;85:11231134.
24. Beck J, Gottfried SB, Navalesi P, Skrobik Y, Comtois N, Rossini M,
Sinderby C. Electrical activity of the diaphragm during pressure support ventilation in acute respiratory failure. Am J Respir Crit Care
Med 2001;164:419424.
25. Delhez L. Electrical responses of the human diaphragm to the electrical
stimulation of the phrenic nerve. Electromyogr Clin Neurophysiol
1975;15:359372.
26. Estenne M, Ninane V, De Troyer A. Triangularis sterni muscle use during eupnea in humans: effect of posture. Respir Physiol 1988;74:151162.
27. Estenne M, Zocchi L, Ward M, Macklem PT. Chest wall motion and expiratory muscle use during phonation in normal humans. J Appl Physiol 1990;68:20752082.
28. Pradhan S, Taly A. Intercostal nerve conduction study in man. J Neurol
Neurosurg Psychiatry 1989;52:763766.
558
29. Whitelaw WA, Markham DR. Electrode for selective recording of electromyograms from intercostal muscles. J Appl Physiol 1989;67:21252128.
30. Whitelaw WA, Feroah T. Patterns of intercostal muscle activity in humans. J Appl Physiol 1989;67:20872094.
31. Saadeh PB, Crisafulli CF, Sosner J, Wolf E. Needle electromyography of
the diaphragm: a new technique. Muscle Nerve 1993;16:1520.
32. Silverman JL, Rodriguez AA. Needle electromyographic evaluation of
the diaphragm. Electromyogr Clin Neurophysiol 1994;34:509511.
33. De Troyer A, Leeper J, McKenzie D, Gandevia SC. Neural drive to the
diaphragm in patients with severe COPD. Am J Respir Crit Care Med
1997;155:13351340.
34. Bolton CF. Clinical neurophysiology of the respiratory system. Muscle
Nerve 1993;16:809818.
35. Gandevia SC, Rothwell JC. Activation of the human diaphragm from
the motor cortex. J Physiol 1987;384:109118.
36. Lindstrm L, Petersen I. Power spectrum analysis of EMG signals and
its applications. In: Desmedt JE, editor. Progress in clinical neurophysiology. Vol. 10. Computer-aided electromyography. Basel, Switzerland: S. Karger; 1983. p. 151.
37. Daube JR. Electrophysiologic studies in the diagnosis and prognosis of
motor neuron diseases. Neurol Clin 1985;3:473493.
38. Gross D, Grassino A, Ross WRD, Macklem PT. Electromyogram pattern of diaphragm fatigue. J Appl Physiol 1979;46:17.
39. Sinderby C, Lindstrm L, Grassino AE. Automatic assessment of electromyogram quality. J Appl Physiol 1995;79:18031815.
40. Daubenspeck JA, Leiter JC, McGovern JF, Knuth SL, Kobylarz EJ.
Diaphragmatic electromyography using a multiple electrode array. J
Appl Physiol 1989;67:15251534.
41. Sinderby CA, Comtois AS, Thomson RG, Grassino AE. Influence of
the bipolar electrode transfer function on the electromyogram power
spectrum. Muscle Nerve 1996;19:290301.
42. Schweitzer TW, Fitzgerald JW, Bowden JA, Lynn-Davies P. Spectral analysis of human diaphragm electromyogram. J Appl Physiol 1979;46:152165.
43. Aldrich TK, Adams JM, Arora NS, Rochester DF. Power spectral analysis of the diaphragm electromyogram. J Appl Physiol 1983;54:15791584.
44. Arvidsson A, Grassino A, Lindstrm L. Automatic selection of uncontaminated electromyogram as applied to respiratory muscle fatigue. J
Appl Physiol 1984;56:568575.
45. Levine S, Gillen J, Weiser P, Gillen M, Kwatny E. Description and validation of an ECG removal procedure for EMGdi power spectrum
analysis. J Appl Physiol 1986;60:10731081.
46. Sinderby C, Friberg S, Comtois N, Grassino A. Chest wall muscle crosstalk in the canine costal diaphragm electromyogram. J Appl Physiol
1996;81:23122327.
47. Basmajian JV, DeLuca CJ. Muscles alive: their functions revealed by
electromyography. Baltimore, MD: Williams & Wilkins; 1985.
48. Desmedt JE. Methodes detude de la fonction neuromusculaire chez
lhomme: myogramme isometrique, electromyogramme dexcitation
et topographie de linnervation terminale. Acta Neurol Psychiatr Belg
1958;58:9771017.
49. Roy SH, DeLuca CJ, Schneider J. Effects of electrode location on myoelectric conduction velocity and median frequency estimates. J Appl
Physiol 1986;61:15101517.
50. Beck J, Sinderby C, Lindstrm L, Grassino A. Diaphragm interference
pattern EMG and compound muscle action potentials: effects of chest
wall configuration. J Appl Physiol 1997;82:520530.
51. Weinberg J, Sinderby C, Sullivan L, Grassino A, Lindstrm L. Evaluation
of diaphragm electromyogram contamination during progressive inspiratory maneuvers in humans. Electromyogr Clin Neurophysiol 1997;37:143.
52. Sinderby C, Lindstrm L, Comtois N, Grassino AE. Effects of diaphragm shortening on the mean action potential conduction velocity.
J Physiol 1996;490:207214.
53. Clarke RSJ, Hellon RFR, Lind AR. The duration of sustained contractions of the human forearm at different muscle temperatures. J Physiol 1958;143:454463.
54. Humphreys PW, Lind AR. The blood flow through active and inactive
muscles of the forearm during sustained handgrip contractions. J
Physiol 1963;166:120135.
55. Edwards RHT, Hill DK, Jones DA. Heat production and chemical
changes during isometric contractions of the human quadriceps muscle. J Physiol 1975;251:303315.
56. Fink R, Luttgau HD. An evaluation of membrane constants and potassium
conductance in metabolically exhausted fibers. J Physiol 1976;263:215239.
57. Roberts DV. Simultaneous measurement of propagation velocity of action potential and contraction wave in frog striated muscle. J Physiol
1969;147:6263.
58. Sinderby C, Navalesi P, Beck J, Skrobik Y, Comtois N, Friberg S, Gott-
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
VOL 166
2002
fried SB, Lindstrm L. Neural control of mechanical ventilation in

respiratory failure. Nat Med 1999;5:14331436.
Parthasarathy S, Jubran A, Tobin MJ. Assessment of neural inspiratory
time in ventilator-supported patients. Am J Respir Crit Care Med
2000;162:546552.
Bellemare F, Grassino AE. Evaluation of human diaphragmatic fatigue.
J Appl Physiol 1982;53:11961206.
Bellemare F, Grassino A. Force reserve of the diaphragm in patients
with chronic obstructive pulmonary disease. J Appl Physiol 1983;55:815.
Sinderby C, Weinberg J, Sullivan L, Lindstrm L, Grassino A. Electromyographical evidence for exercise-induced fatigue in patients with chronic
cervical cord injury or prior polio infection. Spinal Cord 1996;34:594601.
Sinderby C, Spahija J, Beck J. Changes in respiratory effort sensation
over time are linked to the frequency content of diaphragm electrical
activity. Am J Respir Crit Care Med 2001;163:16371641.
Gandevia SC, Plassman BL. Responses in human intercostal and truncal
muscles to motor cortical and spinal stimulation. Respir Physiol 1988;
73:325338.
Lance JW, Drummond PD, Gandevia SC, Morris JG. Harlequin syndrome: the sudden onset of unilateral flushing and sweating. J Neurol
Neurosurg Psychiatr 1988;15:635642.
Linder SH. Functional electrical stimulation to enhance cough in quadriplegia. Chest 1993;103:166169.
DiMarco AF, Altose MD, Cropp A, Durand D. Activation of the inspiratory intercostal muscles by electrical stimulation of the spinal
cord. Am Rev Respir Dis 1987;136:13851390.
Mier A, Brophy C, Estenne M, Moxham J, Green M, DeTroyer A. Action of abdominal muscles on rib cage in humans. J Appl Physiol
1985;58:14381443.
Merton PA, Morton HB. Stimulation of the cerebral cortex in the intact
human subject. Nature 1980;285:227.
Barker AT, Jalinous R, Freeston IL. Noninvasive stimulation of the human motor cortex. Lancet 1985;1:11061107.
Maskill D, Murphy K, Mier A, Owen M, Guz A. Motor cortical representation of the diaphragm in man. J Physiol 1991;443:105121.
Chokroverty S, Hening W, Wright D, Wolczak T, Goldberg J, Burger R,
Belsh J, Patel B, Flynn D, Shah S. Magnetic brain stimulation: safety
studies. Electroencephalogr Clin Neurophysiol 1995;97:3642.
Pascual-Leone A, Houser CM, Reese K, Shotland LI, Grafman J, Sato
S, Valls-Sole J, Brasil-Neto JP, Wassermann EM, Cohen LG. Safety
of rapid-rate transcranial magnetic stimulation in normal volunteers.
Electroencephalogr Clin Neurophysiol 1993;89:120130.
Burke D, Hicks RG, Stephen JP. Corticospinal volleys evoked by anodal
and cathodal stimulation of the human motor cortex. J Physiol 1990;
425:283299.
Burke D, Hicks R, Gandevia SC, Stephen J, Woodforth I, Crawford M.
Direct comparison of corticospinal volleys in human subjects to transcranial magnetic and electrical stimulation. J Physiol 1993;470:383393.
Rothwell JC, Thompson PD, Day BL, Boyd S, Marsden CD. Stimulation of
the human motor cortex through the scalp. Exp Physiol 1991;76:159200.
Baker SN, Olivier E, Lemon RN. Task related variation in corticospinal
output evoked by transcranial magnetic stimulation in the macaque
monkey. J Physiol 1995;488:795801.
Murphy K, Mier A, Adams L, Guz A. Putative cerebral cortical involvement in the ventilatory response to inhaled CO2 in conscious man. J
Physiol 1990;420:118.
Inghilleri M, Berardelli A, Cruccu G, Manfredi M. Silent period evoked
by transcranial stimulation of the human cortex and cervicomedullary
junction. J Physiol 1993;466:521534.
Roick H, von Giesen HJ, Benecke R. On the origin of the postexcitatory
inhibition seen after transcranial magnetic brain stimulation in awake
human subjects. Exp Brain Res 1993;94:489498.
Taylor JL, Butler JE, Allen GM, Gandevia SC. Changes in motor cortical
excitability during human muscle fatigue. J Physiol 1996;490:519528.
Chen R, Collins S, Remtulla H, Parkes A, Bolton CF. Phrenic nerve
conduction study in normal subjects. Muscle Nerve 1995;18:330335.
Moorthy SS, Markand ON, Mahomed Y, Brown JW. Electrophysiologic
evaluation of phrenic nerves in severe respiratory insufficiency requiring mechanical ventilation. Chest 1985;88:211214.
Luo YM, Polkey MI, Johnson LC, Lyall RA, Harris ML, Green M, Moxham J. Diaphragm EMG measured by cervical magnetic and electrical
phrenic nerve stimulation. J Appl Physiol 1998;85:20892099.
Luo YM, Johnson LC, Polkey MI, Harris ML, Lyall RA, Green M, Moxham J. Diaphragm electromyogram measured with unilateral magnetic stimulation. Eur Respir J 1999;13:385390.
Aldrich TK. Transmission fatigue of the rabbit diaphragm. Respir Physiol 1987;69:307319.
4. Tests of Respiratory Muscle Endurance

Muscle endurance is the ability to sustain a specific muscular
task over time. It is a highly integrated and complex quality of
a muscle or a group of muscles that is related to its resistance
to fatigue. To a large extent, any measurement of endurance is
task specific because different tasks result in varying recruitment patterns of motor units and synergistic muscle groups,
each with varying endurance qualities. The wide variety of
techniques that have been developed to measure endurance of
the respiratory muscles differ largely on the type of task that is
being performed. For each specific task, an endurance curve
can be generated by plotting task intensity versus the time it
can be sustained. Task failure is an event defined by the inability to continue performing the required task (Figure 1). At
high levels of intensity, a task can be performed for only a few
repetitions. As the intensity is decreased, each task can be endured for a longer time until a level can be sustained for an indefinite period (i.e., hours). The latter is referred to as the
maximum sustainable task or load. Another estimate of endurance involves performing incremental increases in task intensity for a given time period until a peak intensity is identified, which is the maximum that can be maintained for a finite
period of time (Figure 1). This intensity is not sustainable but
may also be used to reflect endurance properties.
Although respiratory muscle strength and endurance appear
to be closely linked in many conditions (13), there are numerous examples in which endurance would not be accurately predicted from estimates of maximum pressures or maximum ventilatory capacity. Furthermore, the characteristics of endurance
curves for a given muscle may change with training, disuse, drug
treatment, and so on. For example, in heart failure patients (4) or
in normal subjects (5) following certain respiratory muscle training protocols, larger relative effects are seen on endurance compared with strength. Some patients with asthma show inherent
elevations in endurance properties as a fraction of strength (6),
as do patients with cystic fibrosis (7), suggesting these patients
naturally train for endurance during periods of airway obstruction. In contrast, patients with chronic obstructive pulmonary
disease (COPD) (8) or patients receiving acute steroid treatment
(9) show marked reductions in endurance properties relative to
strength. Therefore, endurance measurements can be useful in
some clinical and investigative settings for evaluating patient
populations and responses to treatment and rehabilitation.
MEASURES OF RESPIRATORY MUSCLE ACTIVITY USED

IN ENDURANCE TESTING
Rationale
Many different kinds of tasks have been used to quantify the endurance properties of the respiratory muscles. Most often, endurance has been defined in terms of the ability to sustain a level of
minute ventilation (ventilatory endurance) or a level of inspiratory
and sometimes expiratory pressure. However, these simple measures often present limitations to evaluating the effect of the load
on the respiratory muscles. From a muscle energetics viewpoint,
the energy requirements of a working muscle (and therefore a
rough estimate of its level of activation) are determined largely by
the tension developed over time (i.e., tensiontime product) and
the rate of mechanical work being performed ( W ) (10, 11).

PressureTime Product
Methodology.
Refer to PRESSURE MEASUREMENTS in Section 2
DOI: 10.1164/rccm.AT0402
Internet
address: www.atsjournals.org
of
this Statement
for specific techniques for the measurement
of pressure at the airway opening and esophageal, gastric, and

transdiaphragmatic pressures. The pressuretime product (PTP)
is the integration of respiratory pressure over time (i.e., Pdt).
It is common to express PTP over a 1-minute interval (i.e.,
units pressure time; e.g., cm H2O minutes). The integration process can be performed by most medical amplifiers
or digital computers, much like flow is integrated to obtain
minute ventilation. If such techniques are used, assurances
must be made that expiratory pressures during the expiratory
phase, or inspiratory pressures generated due to chest wall
elastic recoil, are excluded from the analysis of inspiratory
PTP.
A common expression of the PTP is the mean pressure
generated over an entire breath cycle ( P) in Equation 1, in
which
Mean pressure ( P ) = PTP ( sampling period )
(1)
For example, if PTP is measured for a single breath period,

then the sampling period would be total breath period (Ttot).
A signal averaging circuit (available on most medical amplifiers for determining mean vascular pressure) can often be used
to measure P directly, online. These are usually composed of
leaky integrators with time constants of approximately 20
seconds. The analysis can also be done by digital computer or
mechanical devices (12).
The P value calculated in Equation 1 can be measured at
the mouth or airway opening if one wishes to estimate the average pressure generated by all the respiratory muscles working against an external load (i.e., Pmo). Alternatively, it can
be measured using: transpulmonary pressure ( PL) for measurements of activity of the chest wall and its muscles against
the lung and airways (2); transdiaphragmatic pressure ( Pdi)
for the activity of the diaphragm alone (1); or total respiratory
muscle pressure ( Pmus) for activity of the synergic respiratory
muscles against the lung and rib cage (13).
When P is normalized to a fraction of the maximum inspiratory pressure available, it is referred to as the pressure
time index (PTI). For example, for measurements of pressure
at the mouth or airway opening, Equation 2 is
PTImo = Pmo P I ,max
(2)
where PTImo is the pressuretime index measured at the mouth

and PI,max is the maximum inspiratory pressure that can be
generated at the mouth or airway opening (usually obtained for a
period exceeding 1 second). For the PTI for the diaphragm
(PTIdi), maximal transdiaphragmatic pressure (Pdi,max) is substituted for PI,max and Pdi is substituted for Pmo. Refer to
Section 2 of this Statement for techniques of measuring
PI,max and Pdi,max. Traditionally, the term tensiontime index (TTI) has been applied to this measurement (1). From a
physiologic viewpoint, TTI is the ideal variable, which is deterministic for a large number of relevant factors in muscle
physiology, including muscle energetics and blood flow. However, for most experimental and clinical measurements for the
respiratory system, the transduction of muscle tension into
respiratory pressures is not straightforward. Therefore, to
avoid misinterpretation of the data, it is recommended that
PTI be substituted for TTI when pressure comprises the measured variable (see Section 5 of this Statement).
Advantages. Under conditions of relatively constant ventilation, respiratory muscle endurance (1), blood flow (14), and
changes in oxygen consumption of the respiratory system
560
VOL 166
2002
Work Rate of the Respiratory System
Figure 1. A schematic showing an idealized endurance curve for a given

task (e.g., maximal voluntary ventilation) and its relationship to an endurance test (dotted line) as well as to an incremental loading test (thin
stepped line). Tlim Time limit a task can be endured before task failure.
( VO2,rs) (15, 16) have been shown to be significantly correlated to changes in PTI (Figure 2). Furthermore, PTI is a parameter that describes the pressure-generating activity of the
muscles, independent of a specific breathing rhythm, breathing frequency, or type of load within the experimental limits
tested (1). Normalizing to maximum pressure can also be useful as a measure of the amount of pressure reserve utilized
during contraction. For example, most normal subjects can
sustain a PTIdi of up to approximately 0.18 (1) and a PTI for
the chest wall muscles and the synergic inspiratory muscles of
up to approximately 0.3 (2). These critical PTI values may
be useful in estimating whether the muscles are undergoing
contractions that are likely to lead to a loss of force, or fatigue (17, 18). However, critical PTI should be used with considerable caution, as it is highly likely that the critical PTI may
vary somewhat across various pathological conditions. This
has not been studied extensively. In addition, in clinical situations there is often some uncertainty regarding the accuracy of
measurements of PI,max or Pdi,max used to calculate PTI (see
VOLITIONAL TESTS OF RESPIRATORY MUSCLE STRENGTH in Section 2 of this Statement).
Disadvantages. When
the level of ventilation increases at a
constant PTP, the VO2,rs is increased and endurance is reduced (15, 19). For example, in Figure 3A, when a subject is
inspiring with a constant PTP (individual isopleths), increasing
flow rates result in markedly
increased oxygen consumption
of the respirator system ( VO2,rs). Furthermore, when PTP is

kept constant, increasing mechanical work rates of the respira
tory system Wrs result in reduced inspiratory muscle endurance (Figure 3B). Therefore, when the tasks involve high levels of ventilation, as may occur during exercise, during
ventilatory endurance measurements, or in patients with high
or changing ventilatory requirements, the various measures of
pressure over time (i.e., PTP, P, and PTI) become less predictive as global measures of the activity or endurance of the
muscles. Under these conditions, the mechanical work rate
(Wrs), discussed below, begins to take on a greater significance (19). As shown in Figure 3C, when ventilation is al
lowed to vary over
a wide range of PTP, Wrs becomes highly
predictive of the VO2,rs and therefore the energy utilization of

the respiratory muscles.
Another illustration of these points is that the critical PTI
for the respiratory muscles working synergically can vary from
0.12 to 0.4, depending on the particular pattern of ventilation,
particularly when inspiratory flows and timing are varied over
a wide range (13, 20). Nevertheless, under most testing conditions, when ventilation remains relatively low and constant,
and duty cycle is kept within a range that is normally seen during spontaneous ventilation (i.e., 0.30.5), measures of PTP
(alternatively, PTI or P) are still the most predictive global
measure of respiratory muscle activity available.
Methodology. Generally, the ventilatory work rate (power
output) of the respiratory system (Wrs) refers to mechanical

work performed by the respiratory muscles against the lungs
and chest wall during ventilation. It is calculated as the integration of the appropriate measures of pressure volume (see
ASSESSMENT OF THE FUNCTION OF THE ACTIVE CHEST WALL:
CAMPBELL DIAGRAM in Section 6 of this Statement). In this dis
cussion, we will use Wrs to also include the work rate performed by the respiratory system against any external loading
device. Work rate is expressed in joules per minute (1 J 1
kPa 1 L; 1 kPa 10.2 cm H2O). The complete measurement
of work of breathing against the lung and chest wall, for both
inspiration and expiration, is complex, largely because components involving movement and distortion of the chest wall are
difficult to quantify without relatively sophisticated analyses.
However, in many cases, measuring the work performed
against an external load (Wext) may provide sufficient information for purposes of respiratory muscle endurance testing.
If a subject is breathing against an external load and ventilation remains near spontaneous levels during loading, the
rate of work performed against the lung and chest wall remains relatively unchanged from normal breathing. There
fore, any changes in Wrs can be attributed largely to
changes in the work performed against the external load, or
were breathing against an inWext. For example, if a subject
spiratory resistive load, Wext would be directly proportional

to changes in Pmo because (Equation 3)
ext = Pmo V
I ( 15 )
W
(3)
where VI inspiratory minute ventilation. Equation 3 emphasizes one of the reasons why measures of the pressuretime product are so powerful in predicting endurance and
changes in en
ergy consumption during external loading. If VI stays constant,
changes in Pmo become the sole determinant of changes in Wext.
Figure 2. Three important

physiologic variables that are
directly related by the pressuretime index of the diaphragm (PTIdi). (A) Endurance
time (Tlim) of the diaphragm
in human subjects. The critical PTIdi of approximately 0.18
refers to the maximum PTIdi
that is sustainable for a period longer than 2 hours.
Values above 0.18 result in
fatigue and task failure; redrawn by permission from
Reference 1. (B) Diaphragmatic blood flow is affected
by the PTIdi such that above
a critical level, increases in
PTIdi result in reductions in
blood flow; redrawn by permission from Reference 14
(data on dogs). (C) Oxygen
consumption of the respiratory system (VO2,rs) increases
as a function of PTIdi. Measures above 0.2 are difficult
to measure in the steady
state because of fatigue. Redrawn by permission from
Reference 16.
561
Figure 3. (A) Effects of inspiratory flow rate

on the oxygen consumption of the respiratory muscles (VO2,rs) at constant pressure
time products (isopleths). Reprinted by permission from Reference 15 (data on humans). (B) Effects of changes in respiratory
work rate on endurance time (Tlim) at constant pressuretime product. Reprinted by
permission from Reference 19. (C) Effects of
respiratory work rate on VO2,rs when pressuretime product is allowed to vary. Reprinted by permission from Reference 15.
The use of Equation 3 eliminates the necessity of performing complex integrations of individual pressurevolume loops
for each breath, which are required for more sophisticated es
timates of the total Wrs, discussed below. Therefore, it is pos
sible to measure changes in Wext, online, with digital or electronic multiplication of Pmo and VI.
An additional component of Wext occurs from gas compression (expiration) or decompression (inspiration) when large
pressures are generated in the airways (15). During inspiratory
loading, this gas decompression can account for as much as 0.4
L of displaced tidal volume in normal subjects, elevating the
work of breathing by as much as 50%. If thoracic volume is
measured in a volume displacement box, this additional volume
is measured directly. However, it can
also be calculated. Appro
priate equations for adjustment of VI for gas decompression depend on the nature of the loading device (15, 19, 21). For example, if a threshold loading device is used, in which inspiratory
pressure generation is approximately constant, the inspired
minute ventilation can be adjusted appropriately by adding the
decompression volume calculated in the following way:
VT,I = ( FRC + VT,I ) [ 1 ( Pbs Pmo
P H O ) ( Pbs P H O ) ]
2
(4)
where VT,I is the additional inspiratory tidal volume in liters

(due to gas decompression) that must be added for each
breath in the calculation of inspiratory minute ventilation ( VI)

in Equation 3. VT,I is the inspired tidal volume (before gas decompression); FRC is the functional residual capacity in liters
(measured independently); Pbs is body surface pressure (usually atmospheric); PH2O is water vapor pressure at body temperature; and Pmo is the threshold loading pressure at the end
of an inspiration. Of course, all pressures in Equation 4 must
be of the same units (e.g., mm Hg or kPa). For measures of
ventilatory endurance, or when there are changing levels of
ventilation, a significant portion of the work being performed
by the respiratory muscles is done against the resistive and
elastic properties of the lung and chest wall. Therefore, accu
rate estimates of total Wrs must include these measurements.
The work rate against the lung and chest wall is most often obtained by the Campbell method (22), which requires the use of
an esophageal balloon for estimating pleural pressure and
measurement of a relaxationpressurevolume curve for the
lung and chest wall. The original Campbell method (22) is
somewhat tedious to apply practically for routine clinical endurance measurements. Equipment is now available to perform the calculations automatically by computer; but even
with computerized techniques, examination of the breath-bybreath pressurevolume loops is required. For relevant discussions of the appropriate use of the Campbell method and the
Campbell pressurevolume diagram, refer to reviews (2325).
Advantages. As discussed previously, when ventilatory flow
rate increases, total Wrs becomes an increasingly important
determinant of both energy consumption of the muscles and endurance time (Figure 3). For ventilatory endurance testing, mea
surements of Wrs overcome the problems of variability in lung
and chest wall impedance between subjects and in the same subjects over time. Such changes in lung mechanics are inevitable
in patients who may have wide diurnal variations and fluctuations over more extended time periods. Therefore, measurement
of Wrs may be necessary to draw appropriate conclusions regarding the endurance properties in various patient groups. To a
large extent, these studies have yet to be systematically performed.
Whether Wrs or P should be chosen as the primary global

measure of respiratory muscle activity for endurance testing cannot be stated with certainty at this time. It would
be ideal if a
comprehensive relationship between Wrs, P, VO2,rs, and endurance for the respiratory muscles could be derived for all loading
conditions. From an energetics standpoint, the relationship between them is roughly described for the inspiratory muscles by
Equation 5:
O ,rs = W
rs Ers = ( Pmus V
I ) Ers
V
2
(5)
where Ers is the efficiency of the inspiratory muscles and Pmus

is the mean respiratory muscle pressure per breath (15). Equation
5 suggests that if one knew Ers in a given subject, as well as
VI, the energetics and presumably the endurance of the respiratory muscles could be predicted. Unfortunately, Ers is not particularly constant at different relative velocities of muscle shortening (24) or at differing ventilations, depending on the way
breaths are performed (21), making this ideal difficult to obtain.
Disadvantages. The largest disadvantage of monitoring Wrs

during endurance measurements is the complexity of its accurate measurement and analysis. This is not true, however, for the
component of Wrs that comes from Wext. Furthermore, after

decades of studies regarding the work of breathing, there are
portions of chest wall movement and distortion that remain elusive and difficult to quantify under loading conditions. As shown
in Figure 4, distortions of the chest wall are commonly seen as an
adaptive response to external loading (26). Distortions are also
seen during maximum ventilatory maneuvers (27). Finally, the
simple measurement of the relaxation pressurevolume curve
is not easy to obtain in many patients because of the requisite
for complete muscle relaxation (28, 29).
Finally, one component of P that may be important in determining endurance characteristics, and that is not directly re
lated to Wrs, involves the influence of developed pressure on
blood flow during contraction. For example, as Pdi increases,
blood flow to the diaphragm is limited, presumably by the relationships between tissue pressure and vascular conductance
(14, 30) (Figure 2B). Because sustainable task intensities may
in part reflect a balance of energy utilization and supply, it is
likely that the influence of P on muscle perfusion has an independent effect on endurance that cannot be fully accounted for
562
by its mathematical
contribution to Wrs or its energetic contri
bution to VO2,rs.
VENTILATORY ENDURANCE TESTS

Rationale
The goal of ventilatory endurance testing is to define the maximum sustainable ventilation (MSV), usually expressed as a
fraction of maximal voluntary ventilation (MVV). The time
duration needed to define sustainable is a topic of some controversy and varies with the specific technique described below. As shown in Table 1, normal subjects can sustain ventilations ranging from 60 to 80% of MVV. Therefore, with
submaximal exercise, it is probably rare that any normal individuals ever exceed their MSV, because maximum exercise
ventilations average approximately 61 14% of MVV in the
normal population (31). In some athletes, ventilation is maintained near the sustainable level of sedentary subjects. For
example, elite cross-country skiers can sustain ventilation averages during exercise in excess of 100 L/minute, or approximately 61% of their predicted MVV for periods of 30 to 85
minutes (32), with little or no evidence of fatigue. However,
the baseline MVV in these athletes is frequently elevated above
normal, and unlike normal subjects, they can sustain 8690%
of MVV for 4 minutes, presumably because of their extreme
conditioning. In a clinical setting, the measurement of ventilatory endurance takes on a much greater importance because
patients with chronic lung disease or perhaps heart failure (4)
may progress to a condition in which exercise is limited by
their ability to sustain ventilation. The ventilatory endurance
test is a measure of both inspiratory and expiratory muscle endurance.
Methodology
Early techniques for measuring MSV required repeated trials

of MVV with gradually decreasing levels of ventilation, until
an MSV could be determined (33). These have generally been
found to be exhaustive and time-consuming, rendering them
largely impractical for most clinical investigations. However,
more recent methods have been developed that make the procedure more practical to perform, requiring only 1025 minutes/test (4, 34, 35).
For all measurements of MSV in obstructed patients, it is
recommended that the test be preceded by administration of a
nebulized bronchodilator. This may be particularly useful if
ventilatory endurance is to be repeated at different times, for
example, before and after rehabilitation, to reduce inherent
variability in airway resistance.
The test begins with the routine measurement of a 12-second MVV, using the same equipment employed for the MSV
test. Protocols for technique and reproducibility of MVV,
which meet American Thoracic Society (ATS) criteria, are
available (36, 37). Accurate MVV measurements are critically
important for interpretation of MSV. There are two primary
techniques for acquiring MSV, the maximum effort technique
and the maximum incremental technique, as discussed below.
The maximum effort technique requires subjects to target a
ventilation of approximately 7090% of their MVV (7, 34), using visual feedback from a spirometer or an oscilloscope (Figure 5). Sometimes, one or two short practice trials are used to
determine the starting target ventilation. During the first 25
minutes, the target ventilation is adjusted up or down to a
level slightly lower than the subjects maximum effort. The
subject is then continually encouraged to meet the target for the
next 8 minutes. There are some studies that have described
measuring only a 4-minute MVV as an indicator of endurance
VOL 166
2002
Figure 4. Chest wall distortion

during inspiration, shown as rib
cage (solid symbols) and abdominal movement (open symbols),
with progressive increases in inspiratory mechanical loads. The
y axis represents the percentage of tidal volume excursion
in which the abdominal compartment is moving in during
inspiration (abdominal paradox) and the rib cage compartment is distorted out during inspiration, along a path that is
greater than its expected path of expansion. Reprinted by permission
from Reference 26.
(32). Although a potentially useful and practical approach, insufficient data are available to evaluate whether this provides
a sufficient estimate of sustainable ventilation. In all studies, it
is necessary to control end-tidal carbon dioxide (PETCO2) during the test, usually by adjustment of the carbon dioxide fraction (FCO2) in the rebreathing dead space. The average ventilation achieved over the last minute is considered
to be the
MSV. It is not routine to measure the Wrs or VO2,rs, but there

are rational advantages in doing so, as discussed below.
There is no standardized equipment available for measuring ventilatory endurance. However, the system used should
have the following capabilities: (1) provide for maintaining
isocapnia during hyperpneic maneuvers; (2) have a low impedance to air flow, which meets accepted standards for
spirometry such as ATS criteria (e.g., 2.5 cm H2O/L/second
to 14 L/second) (36); (3) provide reasonable humidification of
the inspired air; and (4) provide real-time visual feedback of
ventilation. Mechanical systems that approach these criteria
have been described in the literature (5, 7, 34, 38), one of
which is illustrated in Figure 5. Care must be taken if pneumotachographs are used for ventilation measurements, to ensure
that they are linear over the range of measured flows, that
their electronic drift is compensated for, and that they do not
contribute significantly to the resistance of the system. If the
pneumotachograph is in the patient line, appropriate compensations should be made for changes in gas viscosity due to supplemental oxygen if it is used.
The maximum incremental technique is a newer procedure
for obtaining an estimate of MSV. It uses 10% incremental increases in target ventilation, every 3 minutes, beginning at
20% of MVV until the subject cannot sustain the target ventilation for the last 3-minute period (4, 35) (MSV is calculated
from the last 10 breaths of the last minute of the highest target
ventilation). This technique, which resembles an incremental
exercise test, was demonstrated to result in MSV measure-
TABLE 1. PREDICTED VALUES FOR MAXIMUM SUSTAINABLE

VENTILATION/MAXIMUM VOLUNTARY VENTILATION
Author (Ref.)
Keens (7)
Keens (7)
Leith and Bradley (5)
Bai (39)
Belman and
Gaesser (69)
Mancini (4)
Subject Age
(yr)
26 2
31 1
31 3
31 3
67 4
(estimate)
50 1
No. of Subjects
Subject Sex
(No. F/M)
MSV/MVV*
(%)
16
14
12
5
25
16/0
0/14
1/11
0/5
14/9
60 8
62 9
80 6
75 4
63 11
1/7
55 9
Definition of abbreviations: F female; M male; MSV maximum sustainable ventilation; MVV maximum voluntary ventilation.
* Results represent means SD.
563
ments nearly identical to those that could be attained by traditional approaches, and was well tolerated by subjects (4).
The importance of sustaining a maximum ventilation for the
three or more minutes at the end of the test should be emphasized. Presumably, during this period, fatigue of the respiratory muscles is progressing rapidly because it is a period of
maximum effort following a relatively long period of nearmaximum effort. Presumably this results in a decay of ventilation to a near sustainable level.
Normal Values
Normal values for MSV, by any method, have not been systematically obtained over a wide population, and results vary
considerably between laboratories (Table 1). The large differences in predicted values may be due in part to variations in
technique, particularly with respect to impedances of the mechanical measuring devices. The system impedance can have
substantial effects on the total Wrs at high ventilations. In addition, there are differences in the populations studied and
what was defined as sustainable. It is recommended that when
publishing reports of ventilatory endurance, the value for the
impedance of the measuring device be stated. Until more
complete population standards and uniform equipment and
techniques are available, each laboratory is advised to establish its own population standards.
Results for MSV should be reported as a fraction of measured MVV (MSV/MVV%) and either as an absolute value
(L/minute) or as a fraction of predicted MVV (MSV/MVV%
pred). The latter, which has not been used routinely, provides
a normalization of the absolute sustainable value to the patients age, height, and sex, independent of inherent lung or
respiratory muscle function.
Advantages
There are a number of advantages to measuring MSV as an indicator of respiratory muscle endurance, the most important
of which is its close resemblance to the task performed during
exercise. It therefore provides clinically relevant data that can
be related to function. Second, it is probably a measure of
both inspiratory and expiratory muscle endurance because in
normal subjects there appear to be decrements in both inspiratory and expiratory function after MSV maneuvers (39).
Finally, maximum ventilatory maneuvers result in evidence of
diaphragm fatigue (3941). Interestingly, this does not appear

to be true for patients with COPD (42).
Disadvantages
The disadvantages of using MSV as an indicator of endurance

are related to the difficulty in estimating the relative contribution of lung and chest wall mechanics to the measurement.
MVV measurements are highly susceptible to relatively small
changes in flow resistance, the effects of which are amplified
exponentially as ventilation increases (24). Therefore, the load
on the respiratory muscles is not uniform across patients or
even in the same patients over time.
This is of considerable importance in COPD or other obstructive lung diseases in which day-to-day and diurnal variations
in airway mechanics are common. Second, the wide variety of
strategies utilized in a given patient to perform MVV-like maneuvers leaves many potential sources of variance between
subjects. For example, in patients with COPD, effective use of
the expiratory muscles is often limited during elevated ventilations (compared with normal subjects) because of early maximum flow limitation. This is accompanied by hyperinflation
and shorter inspiratory muscle lengths with a greater proportional burden on the inspiratory muscles than would be seen
in normal subjects. In patients with COPD, measurements of
MSV/MVV% as an indicator of respiratory muscle endurance
have suggested excellent ventilatory muscle endurance relative to strength, as compared with control subjects (34). However, because their mechanical abnormalities have greater relative influence at higher ventilations, the denominator of the
MSV/MVV fraction may be artificially low in these patients,
giving the impression that endurance properties are normal.
When external resistive loading techniques are utilized, which
reduce the contribution of lung and chest wall mechanics as a
factor in the measurement, it is found that the endurance capacity of the respiratory system is relatively low in the COPD
population compared with normal subjects (8).
The problem of the contribution of the inherent impedance
of the respiratory system could be overcome by careful mea
surement of Wrs during the test. As shown in Figure 6, redrawn from the experiments of Tenney and Reese (33), a
strikingly different view of endurance can be seen when work
rate or power output is quantified. Although this subject could
sustain approximately 68% of his MVV, he could sustain only
approximately 30% of maximum Wrs. The investigators also

showed that despite experimental alterations in pulmonary
Figure 5. A typical apparatus for measuring ventilatory endurance.

564
VOL 166
2002
impedance, the Wrs-versus-Tlim relationship did not change

appreciably (33).
In summary, ventilatory endurance testing can be useful as
a functional measurement, particularly in the setting of rehabilitation or other forms of treatment. Results should be
viewed with some understanding of the tests limitations with
respect to separation of muscle properties versus intrinsic mechanical properties of the lungs and chest wall. This limitation
could potentially be overcome by measuring Wrs, although

this may not be practical in many clinical settings. The most
promising approach appears to be the incremental method.
However, full support of the technique awaits verification in
other laboratories.
ENDURANCE TO EXTERNAL LOADS

Rationale
When an external mechanical load is applied to the airway

opening, the respiratory muscles must generate an additional
pressure to overcome the impedance of the load. The external
load can be one of several types: (1) a flow resistive load, in
which the pressure required of the muscles is dependent on
the flow rate across the resistance. Flow resistive loads can be
linear or nonlinear depending on whether they produce laminar or turbulent flow; (2) elastic loads, in which the pressure
required of the muscles is dependent on lung volume. The
higher the tidal volume, the higher the pressure required.
Such loads are flow independent; (3) threshold loads, in which
a finite pressure is required to open a valve that allows flow to
occur. Therefore, the pressure required of the muscles at the
airway opening is relatively constant, independent of both volume and flow. Threshold loads result in contractions that are
similar to isotonic contractions; or (4) isoflow loads, in which
the flow rate and therefore, the rate of inflation is held constant and the pressure generated against the flow is a measured output variable. Isoflow loads are similar in concept to
isokinetic contractions of limb muscles, in which velocity of
shortening is held constant.
To conduct a respiratory muscle endurance test with an external load requires setting the task that the subject must perform against the load. For example, the subject may be asked
to breathe normally or to breathe with a set breathing pattern
or with a specific muscle configuration. Different ways of contracting against the load result in markedly different measures
of endurance, reemphasizing the importance of the concept of
task specificity.
The advantage of using externally applied loads is that it is
much easier to control the relevant variables during the test. It
is even possible to design tests that are specific to the diaphragm (1) or the rib cage muscles (2, 43). Generally, these
tests require large developed pressures against normal or relatively modest changes in ventilatory requirements. Such conditions are similar to those of weight lifting, with relatively low
velocities of shortening. In contrast, measures of ventilatory
endurance, described previously, are more like activities of
running with large velocities of shortening and participation
by a large number of synergic muscle groups. Interestingly,
measurements of endurance to high inspiratory resistive loads
appear to be more a reflection of rib cage muscle endurance
than diaphragm endurance (44). Therefore, the exact extent
to which measurements of endurance to high external loads
apply to ventilatory endurance or to clinically relevant conditions such as exercise has not been well defined.
A large number of devices and techniques have been developed to measure endurance of external loads. The most common is the use of orifice-type flow resistance applied to the in-
Figure 6. Relationships between power output, ventilation, and endurance time (Tlim). Solid line: Maximum voluntary ventilation. Dotted
line: Calculated work rate or power output of the respiratory muscles.
Redrawn by permission from Reference 33.
spiratory circuit (20, 45). Excellent studies can be performed

with flow resistances, but because the pressure load seen by
the respiratory muscles depends on the developed flow, the
technique requires visual feedback of some form of ventilation, preferably the flow rate. Therefore, for practical reasons,
flow resistances have largely been replaced in most clinical
laboratories by threshold loading devices or other techniques
discussed below. The techniques below have generally been
used to measure inspiratory muscle endurance.
Maximum Sustainable Threshold Loading
Methodology. Nickerson and Keens (46) developed a method

in which endurance times are measured in response to gradually decreasing threshold pressures, starting near PI,max. They
described one of the first threshold loading devices that was
relatively flow independent. The test usually begins with a careful
measurement of PI,max. Sequential Tlim measurements are
then made, beginning at approximately 90% of PI,max and
decreasing in increments of 5%. Subjects are allowed to rest
between each measurement for approximately 10 times the
length of Tlim. No attempt is made to control the breathing
pattern. Task failure is determined at each load by the inability to maintain ventilation against the load, resulting in the
subject coming off the mouthpiece. Other definitions of task
failure define a point at which a subject is unable to generate
the threshold pressure or a target flow for three consecutive
breaths (47). The first pressure that can be sustained for more
than 10 minutes is considered the sustainable inspiratory pressure (SIP). The SIP is determined by averaging the pressures
over the last 20 breaths.
The original Nickerson and Keens (46) threshold loading
device has never been available commercially, but is made of a
simple plunger, with leaded rings added to the inside of the
chamber to weight the valve. A more modern version is illustrated in Figure 7. There is a linear relationship between increases in weight and the pressure required to lift the plunger.
The original device used a plunger, seated with a 1-in. O ring
onto a 45 surface (46). Larger O rings result in more flow independence but less stability. Even small changes in the size of
the contact circumference and the precision of the seating can
have large effects on the weight/pressure relationship. Therefore, each homemade valve requires independent additional
supports for the plunger, which improve its stability (48), and
the use of standardized, commercially available valve mechanisms (nondisposable positive end-expiratory pressure valves),
which improve the pressureflow characteristics (Figure 7)

(49). Some commercially available spring-loaded threshold
valves do not have the pressure range necessary for testing endurance in most patients.
Normal values. As with most respiratory muscle endurance testing techniques, normal values have not yet been developed. For example, the influence of stature, age, and sex is
not described and the numbers of subjects have been low.
Nickerson and Keens (46) tested 15 normal individuals ranging from 5 to 75 years of age. The 12 adults could maintain a
mean SD SIP of 82 22 cm H2O, or 71 10% of PI,max.
On a second trial, in 12 subjects, both PI,max and SIP increased by approximately 10%, while the relationship of SIP/
PI,max remained constant. Somewhat different results were
found by Martyn and coworkers (50) when using the method
of Nickerson and Keens (46). They found that the SIP was 52
6% of PI,max on the first trial. However, when subjects were
asked to repeat the loads that they had previously failed, they
were able to increase their SIP to 77 6% of PI,max (50).
Advantages. The attraction of the technique of Nickerson
and Keens (46) has been that it provides a method for evaluating
global respiratory muscle endurance in a one-session test,
much like a pulmonary function test. There were no previous
studies that defined a technique to establish sustainable pressure
in a practical setting. Furthermore, the test is noninvasive and
is tolerated relatively well, the equipment required is inexpensive
and does not require a great deal of training or coordination
for the subject, and the results are relatively independent of the
mechanics of breathing because minute ventilation increases
minimally.
Disadvantages. It is clear that subjects will adjust their
breathing pattern as they attempt to breathe against any kind
of large mechanical load, and they will learn to do this over
time (51). This effect may have been underestimated by Nickerson and Keens (46) as discussed by Martyn and coworkers
(50). Relatively small changes in duty cycle (52), inspiratory
flow rate (20, 52), and tidal volume (13) can have relatively
large effects on endurance measures. Therefore, it would seem
appropriate to control the pattern of contraction against the
load during the test. However, it is likely that a naive subject
will be able to achieve longer Tlim values when allowed to
breathe spontaneously. Artificially imposing a breathing pattern may not be appropriate for body size, vital capacity, or
CO2 production. Furthermore, chest wall configuration, and
therefore respiratory muscle recruitment, are quite different
when inspiring against target respiratory patterns, when agonists and antagonists are simultaneously recruited (52), as
compared with spontaneous or maximum uncontrolled inspirations (13). Nevertheless, the effects of the pattern of contraction and recruitment on Tlim result in an inherent measurement variability between subjects and in the same subject
over time (50). It is likely that this problem could be overcome
to some extent by measuring Pmo and Wext, because they are

likely to be the most dominant determinants of Tlim, regardless of the pattern of breathing. However, this has not been
measured systematically in available clinical studies using
maximum sustainable threshold loading.
Having subjects begin with endurance trials at the highest
pressure loads can be exhausting, uncomfortable, and timeconsuming for the patient. The test generally requires a minimum of 2 hours, as was originally described (46).
Maximum Incremental Threshold Loading
Methodology. The incremental threshold loading technique

was described in the late 1980s (8, 50, 53). In concept, it was
designed to resemble a Bruce protocol, which is popular for
565
Figure 7. (A) An example of a threshold loading device built from a

commercially available positive end-expiratory pressure valve (Ambu
PEEP valve; Life-Assist, Rancho Cordova, CA). (B) Typical pressureflow
characteristics of the device, showing the relative flow independence
of the threshold pressure. Reprinted by permission from Reference 49.
incremental, whole body exercise testing. Before the study,

the patients PI,max is measured by standard techniques (see
PRESSURE MEASUREMENTS in Section 2 of this Statement). The
subjects inspire from a threshold valve, as described previously, beginning with initial threshold pressures of approximately 3040% of PI,max. The threshold pressure is then increased by a unit of weight (e.g., 100 g) added to the outside of
the valve, resulting in a change in pressure of approximately
510% of PI,max, until the load cannot be tolerated for 2 minutes. The maximum inspiratory mouth pressure that can be
tolerated for the full 2-minute interval is considered the peak
pressure (Ppeak) (8). This technique has generally been applied to the measure of inspiratory muscle endurance.
Normal values. Normal values obtained by this technique
have not been thoroughly described. Small sets are displayed
in Table 2. Results are relatively consistent between laboratories, particularly for the peak PTI that can be achieved in the
last stage (PTIpeak, 0.25 to 0.32). Older subjects appear to
demonstrate less initial strength (PI,max) but an ability to
achieve higher relative PTIpeak values (8), which may be consistent with chest wall adaptations to aging (54).
Advantages. The incremental threshold test holds strong
appeal as a measure of inspiratory muscle function because it
is well tolerated and provides a clear outcome variable that is
somewhat easier to define than sustainable efforts. Furthermore, it appears to be sensitive to disease states and clinical
treatment (3, 8, 9, 55). The test has been described by the original authors as more reproducible than the technique of Nickerson and Keens (46); as being tolerated well by naive subjects, who give results similar to those of trained subjects; and
566
having an overall outcome that is little affected by the breathing pattern (56). However, subsequent testing by Eastwood
and coworkers (57), using similar techniques, found that normal subjects demonstrate a considerable learning effect from
the first to the third trial.
An interesting observation is that a peak Wext reaches its

highest value and then falls precipitously before attaining
Ppeak, while oxygen consumption and pressure development
are still rising (50). This means that efficiency is falling during
the final stages of the test. It is promising that peak Wext may
be a useful measurement of the capacity of the muscles of the
chest wall to perform external work, a value that may be as
important clinically as the measure of endurance. A second interesting finding is that some subjects are able to achieve
higher Ppeak values during this test than they can attain during PI,max maneuvers (50, 53, 57), a phenomenon attributed
to the fact that some subjects are unable to maximally activate
their inspiratory muscles during PI,max testing (58).
Disadvantages. Unfortunately, the extent to which the results from incremental tests represent endurance or strength is
not entirely clear. Strictly speaking, it is not a test that has been
proven to be a direct measure of endurance, just as an incremental exercise test is not generally considered an endurance
test. However, when individuals are asked to attempt to sustain
the maximum threshold pressure previously reached with the
incremental method, the average Tlim they can maintain is approximately 6 minutes (53). This suggests that although the
maximum incremental threshold pressure is not sustainable, it
is certainly approaching the asymptote of a typical inspiratory
muscle endurance curve. Another suggestion that the threshold
test is approximating an endurance measurement is the fact that
the maximum PTImo in the last stage is approximately 0.22
0.32 (50, 57), which is similar to sustainable pressures described
for the rib cage muscles (2) and the inspiratory muscles working
in synergy in a normal range of duty cycles and low flow (13).
Of some concern is the tendency for hypoventilation and
desaturation during the test (57). Although modest desaturation is unlikely to affect the measurement appreciably in normal subjects (57, 59), hypercapnia may contribute to a loss of
function, unrelated to endurance characteristics (60, 61). Finally, as the intensity of the load increases, subjects consistently decrease end-expiratory lung volume to maximize the
length and configuration of their inspiratory muscles (57).
This is something of a disadvantage for testing, because the capacity of the muscles to contract against the load is changing
during the test. However, it is likely that such changes in configuration are typical of patient responses to many types of
high inspiratory mechanical or ventilatory loads and is not a
problem unique to incremental loading. Finally, the recruitment patterns of the respiratory muscles may vary during incremental loading and may not totally reflect the endurance
characteristics of breathing against constant submaximal loads.
Repeated Maximum Inspiratory Pressures
Methodology. McKenzie and Gandevia (6, 62, 63) have developed a technique that uses 18 repeated PI,max maneuvers.
The test begins with measurement of PI,max and practice efforts using visual feedback of airway opening pressure. Three
different breathing patterns have been described (6, 62). The
most practical appears to be a series of 18 PI,max contractions
lasting 10 seconds each, with 5 seconds of rest between contractions (duty cycle 0.67) (62). A similar approach has
been used to measure expiratory and limb muscle endurance
(6). The only equipment required is a manometer for measuring airway opening pressure. This technique has been generally used to measure inspiratory muscle endurance.
VOL 166
2002
Normal values. In normal young subjects (n 12), with a

duty cycle of 0.67, the average inspiratory mouth pressure attained in the last contraction is 87 3% of PI,max (mean
SD) (6). The PTImo at this point is approximately 0.58. Using
a similar protocol, but with a slower frequency, pressures
dropped to approximately 77% (6). Interestingly, when the
duty cycle is reduced to 0.5, no drop in pressure generation is
observed across the 18 contractions in normal subjects (6).
The PTImo is then 0.25, which may be just below the threshold for fatigue for the rib cage muscles (2, 43).
Advantages. The technique provides a measurement that
is entirely independent of lung and chest wall mechanics, as
well as mechanical work of breathing, making it potentially
useful for understanding endurance properties of the respiratory muscles without interference from chest wall or lung mechanics. It appears to be sensitive to the influence of lung disease (6), is simple to perform, and lends itself to the potential
for a pulmonary function testing environment.
Disadvantages. Potential disadvantages include the fact that
the endurance characteristics may reflect the anaerobic capacity
of the muscles to sustain force, rather than aerobic endurance,
because it is likely that blood flow is largely occluded to the
muscles during the prolonged contractions. It also does not appear that in 18 contractions a sustainable level of pressure is
fully attained (62). Furthermore, patients with severe lung disease may find it difficult to perform such extended maximum
inspiratory maneuvers without discomfort or dyspnea. This technique has yet to be independently tested in patients.
Maximum Sustainable Isoflow
Methodology. The isoflow method allows subjects to inspire

with PI,max against a device that provides a constant inspiratory flow rate to the lungs (14, 64) (Figure 8). In this way, it resembles the repeated PI,max technique but the lungs are inflated and the inspiratory muscles are allowed to shorten at a
relatively constant rate. The method was modeled after isokinetic testing devices commonly used in limb muscle evaluation. Visual feedback of inspiratory pressure, over time, is provided from an oscilloscope. Breathing pattern is generally set
such that the subjects hyperventilate during the test. The inspiratory airflow is humidified, and PETCO2 is maintained at
eucapnia with supplemental CO2. For routine measurements,
inspiratory flow is maintained, at approximately 1 L/second,
inspiratory time at 1.5 seconds and total breath period at 3.5
seconds (duty cycle 0.42). Many other breathing patterns
have been used with this technique (14, 64, 65); however, for
normal subjects this pattern has been shown to be well tolerated. Subjects continue to inspire maximally with each breath
for 10 minutes. Airway opening pressures generally decline
exponentially during this period until a sustainable pressure
is obtained (Figure 8). Using curve-fitting techniques, it has
been shown that sustainable pressures in normal subjects can
be calculated within 5% with only 5 minutes of endurance
testing (64).
To roughly calculate the additional inspiratory pressure used
to overcome lung and chest wall impedance, the isoflow apparatus can be modified to inflate the subjects lungs during
complete relaxation (13). This additional positive pressure can
be added to active inspiratory pressures developed during each
breath to estimate the total inspiratory muscle pressure (Pmus).
The isoflow apparatus consists of a large and well-regulated pressure source providing inspiratory flow across an extremely high resistance (13, 64). The pressure drop across the
resistance is so high (8,000 to 14,000 cm H2O) that any additional inspiratory pressures developed by the subject at the
airway opening have negligible effects on flow rate. Flow is
567

TABLE 2. PREDICTED VALUES FOR INCREMENTAL THRESHOLD LOADING*
Ppeak/PI,max
(%)
Author (Ref.)
Age
(yr)
Subjects No.
(M/F)
PTIpeak
Notes on End Point
Martyn (50)
33 2
14 (9/5)
88 10
NR
McElvaney (53)
31 5
10 (5/5)
0.22 0.07
0.25 0.08
Morrison (56)
67 4
8 (5/3)
Trial 1:
84 17
Trial 2:
87 21
80 17
Ppeak determined from highest

load over 1 min
Highest Ppeak tolerated for
full 2 min
Eastwood (57)
30 (2841)
7 (5/2)
Trial 1:
75 20
Trial 3:
94 21
0.32 0.12
NR
0.26 0.11
Highest Ppeak tolerated for

full 2 min
Highest Ppeak tolerated for 30 s
Definition of abbreviations: F female; M male; NR not reported; PI,max maximum inspiratory pressure; Ppeak peak threshold
pressure achieved during incremental loading under conditions stated in Notes on End Point column; PTIpeak peak pressuretime index
achieved during incremental loading under conditions stated in Notes on End Point column.
* Results represent means SD.
PI,max was measured at residual volume.
Mean (range).
initiated by negative mouth pressures of 2 to 3 cm H2O

and turned off at 2 to 3 cm H2O by an electrical triggering
circuit. Subjects are protected from the high-pressure source
by breathing from a nonrebreathing valve, which will ensure
that flow bypasses the mouth if there is no active inspiration.
End-tidal CO2 is monitored continuously, and additional CO2
is bled into the inspiratory line to maintain PETCO2 at eucapnia.
This technique has been used primarily to measure inspiratory
muscle endurance.
Normal values. Normal values have not been well described over a wide range of subjects. However, in 15 normal
subjects (8 males and 7 females; age, 26 6 years) with
breathing patterns described above, the peak airway pressure
dropped to 70 7% of their initial pressures (measured with
inspiratory flow of 1 L/second), and 61 12% of PI,max by
the end of 10 minutes of repeated contractions (64). The sustainable PTI with the pattern of contraction described above
was 0.18 0.04. There is a small but significant training effect
between the first and fourth trials with the procedure (64).
Advantages. The isoflow technique has the advantage that
most of the important parameters influencing respiratory endurance measurements are controlled. For example, PETCO2
(and therefore arterial oxygen saturation), breath timing, inspiratory flow, and tidal volume are fixed. Furthermore, lung
and chest wall mechanics can be accounted for at a first approximation (13). An additional strength is the fact that it is
possible to measure the inspiratory muscle strength under
similar conditions used in the endurance test. This avoids the
difficulty of comparing pressure measurements under static
contractions (PI,max) with contractions under dynamic conditions, where changes in length and velocity of contraction affect pressure development (13, 64). Furthermore, because subjects are performing maximal contractions, the fatigue process
develops rapidly and the sustainable pressures can be obtained in
a few minutes of testing. The decay of inspiratory pressure
over time is an additional variable that can be helpful in distinguishing effects on the fatigue process, independent of sustainable pressure development (61). The test is noninvasive and is
tolerated well by naive subjects.
Disadvantages. As yet, the isoflow technique has not been
used to test patient populations and therefore its utility has
not been determined in the clinical setting. It has, however,
been shown to be useful for studying mechanisms of fatigue
(13, 59, 61, 65). One potential problem with applying the tech-
nique on patients is the difficulty with imposing the same

breathing pattern used on normal subjects. For example, normal subjects have relatively high ventilatory requirements during
the test to assure maintenance of ETCO2, whereas patients with
lung disease may not be able to physically perform such high
levels of ventilation. Furthermore, the method also depends on
subject cooperation, and one cannot be certain of the relative
contributions of the rib cage or the diaphragm during contractions. Finally, the equipment used for the isoflow technique is
not available commercially, although it is not particularly ex-
Figure 8. (A) Equipment used for the isoflow loading device. The tanks
and regulators on the left provide a high-pressure source through an
extremely high resistance. Flow is activated by a pressure-triggering
device. The oscilloscope provides visual feedback, so that the subjects
can perform maximally. CO2 is maintained at a constant value by supplementing the inspiratory gas. (B) A typical endurance curve from an
isoflow test. Open circles: The peak pressure developed during inspiration. Open triangles: The average pressure generated during inspiration
per breath. Reprinted by permission from Reference 64.
568
pensive to create from basic equipment in most physiology

laboratories.
In summary, at the present time, for clinical applications,
the most promising and practical technique for evaluating the
endurance qualities of the global inspiratory muscles against
external loads appears to be the incremental threshold loading
technique, originally described by Martyn and coworkers (50)
and refined by later studies (8, 53, 56, 57). However, because
of the uncertainties with regard to its specificity for endurance, we recommend that the term maximum incremental
performance be used to describe its outcome measures until
further information is available. The usefulness of the tech
nique will be advanced by careful quantification of Pmo, Wrs,
and VO2,rs during the tests and by consistent recording of the
maximum values that can be maintained for the full 2-minute
increments. Further work needs to be done to define predicted values in the normal population.
Other methods described here are of considerable value
under experimental conditions and should be considered as
options, particularly for specific experimental designs where
careful control of the variables is critical. Whenever possible,
with any of these techniques, the goal should be to define the
sustainable level of P and Wrs. Further studies in patient populations will be required to determine the comparative usefulness of these techniques in a clinical environment.
All of the external loading techniques are more likely to reflect the endurance qualities of the rib cage muscles as compared with the diaphragm (44, 57, 66). This should be kept in
mind with regard to their clinical implications.
VOL 166
2002
above 0.18 were sustainable for shorter periods. This work

proposed the concept of PTIdi as a consistent index to predict
development of fatigue and failure. Accomplishing this requires active contraction of the abdominal muscles during inspiration, resulting in the diaphragm generating approximately
equal but opposite pressures against the abdomen and rib
cage. The choice of using a Pga of 50% of Pdi was arbitrary
and may not have been a critical variable. A procedure for
quickly determining a sustainable pressure load has not been
developed. Bellemare and Grassino (1, 67) have used this
technique primarily to understand the determinants of diaphragm fatigue by performing repeated endurance trials with
varying Pdi targets and did not intend its use as a clinical test.
The equipment required for this technique is shown in Figure 9. It is possible that the use of a threshold resistance may
be of benefit for controlling esophageal pressure (Pes) during
inspirations rather than an orifice-type resistance, thus eliminating the need for controlling both Pdi and abdominal pressure (Pab) with visual feedback.
Normal Values
Measurements of diaphragm endurance have not been routinely performed in a large number of normal subjects. However, Bellemare and Grassino (1) described the normal sustainable PTIdi to be in the range of 0.150.18 when tidal
volume remained approximately 0.75 L (Figure 2A). Therefore, at a duty cycle of approximately 0.4, normal subjects can
sustain approximately 4050% of Pdi,max.
Advantages
ENDURANCE OF THE DIAPHRAGM

To specifically load the diaphragm requires the subject to attempt to generate a target transdiaphragmatic pressure rather
than a target mouth pressure. This is because it is possible to
generate inspiratory pressures at the airway opening by the rib
cage muscles in the absence of significant diaphragm contribution. Furthermore, pressure development by the diaphragm is
not only distributed against the rib cage but is used to contract
against the abdominal contents as well.
The technique described by Bellemare and Grassino (1) is the

only method that has been identified for specifically loading
the diaphragm and ensuring that it fatigues during endurance
measurements. By limiting tidal volume and contracting the
abdominal muscles (i.e., diaphragm antagonists), changes in
diaphragm shortening with contraction are reduced to a minimum. This means that the energetics of diaphragm contrac
tion are uniquely dependent on Pdi and not Wdi because the
diaphragm is doing little mechanical work against the rib cage
or abdomen.
Methodology
Disadvantages
Roussos and coworkers (45) tested for the maximal sustainable, transdiaphragmatic load. Subjects sustained a given Pdi
until they could no longer reach the target pressure. There
were no requirements on breathing frequency or duty cycle.
They found that approximately 40% of Pdi,max could be sustained for 6090 minutes. A higher Pdi lasted for a shorter
time.
A more precise technique for measurement of diaphragm endurance in humans was developed by Bellemare and Grassino
(1, 67). Subjects are instrumented with an esophageal and gastric
balloon as described in PRESSURE MEASUREMENTS in Section 2 of
this Statement. Maximum transdiaphragmatic pressure is determined and then subjects proceed to inspire through a variable inspiratory flow resistance with a set breathing pattern by
watching an oscilloscope (Figure 9). Two target pressures are
displayed on the oscilloscope screen: Pdi and gastric pressure
(Pga). Tidal volume and duty cycle (TI/Ttot) are monitored as
well. The subject generates a target Pdi by actively inspiring
against a variable resistance and a target Pga of approximately
50% of Pdi. Runs with Pdi of 0.2 to 0.8 and TI/Ttot of 0.2 to 0.7
were tested. The product of Pdi/Pdi,max TI/Ttot was found
to be the best predictor for endurance. Values of 0.15 to 0.18
or smaller could be sustained for more than 1 hour. Values
Because of the need for invasive instrumentation and the requirement of the subject to coordinate a rather unnatural pattern of abdominal and thoracic expansion, the technique has not
been applied extensively to the clinical environment. PTIdi
was measured in patients being weaned from a ventilator. It
shows that patients developing a PTIdi higher than 0.18 failed
the weaning trial. The same patients were tested at a later date
with favorable evolution and had a PTIdi below 0.15, and they
could be weaned (68). One potential complication lies in the
fact that diaphragm blood flow may be determined in part by
the relative negative or positive pressures on its surface. For
example, Buchler and coworkers (30) demonstrated that blood
flow is obstructed more by high positive abdominal pressures
than by similar Pdi values obtained by negative pleural pressures. This would suggest that by contracting the abdominal
antagonist muscles simultaneously with inspiration, there may
be reduced blood flow to the diaphragm, resulting in a greater
probability of fatigue and a lower endurance.
In summary, for a specific measure of diaphragm endurance, the technique of Bellemare and Grassino (1) remains the
only method currently available. The methodology may become more accessible in the clinical environment as techniques develop for rapidly attaining a measure of sustainable
Rationale
569
Figure 9. The method for measuring diaphragm endurance. Pdi

Transdiaphragmatic pressure; Pdi,max maximum transdiaphragmatic pressure; Pga gastric pressure; V respiratory flow from a
pneumotachograph; PETCO2 end-tidal carbon dioxide pressure; VT
tidal volume. Redrawn by permission from Reference 1.
PTIdi, using incremental or maximum effort approaches similar to those described for global inspiratory muscle function.
This technique remains in the domain of clinical research. Few
studies are available.
CONCLUSION
This Section of the Statement has explored the available techniques to assess respiratory muscle endurance. The measurements and techniques include the following:
1. Pressuretime product (PTP): The integration of inspiratory pressure swing over time. Pressure can be esophageal,
Pdi, or mouth pressure (if an external resistance is added to
the circuit). PTP of the expiratory muscles can also be measured. If pressure is normalized to a fraction of the maximum pressure, the value obtained is the pressure time index (PTI). A PTI of 0.150.18 is the upper limit that can be
sustained indefinitely by the diaphragm in humans and as
high as 0.3 for the rib cage muscles. The PTI thresholds are
lower if inspiratory flow is high.
2. Work of breathing: Calculated by the integration of pressure on tidal volume, measures work against an external inspiratory or expiratory load and is a useful test for measuring endurance as a function of workload. Values of work of
breathing relate well to oxygen consumption over a wide
range of ventilations. This measurement is limited to respiratory research and could benefit from computerized equipment to facilitate measurement and analysis in the clinical
setting.
3. Ventilatory endurance tests: Maximal sustainable ventilation (MSV) expressed as a percentage of 12 seconds of
maximum voluntary ventilation. Two techniques are available to determine MSV: the maximum effort technique
(the subject seeks to sustain ventilation at a target level of
7090% MVV for 8 minutes) and the maximum incremental technique (starting at 20% MVV, the target ventilation
is increased by 10% every 3 minutes). There are limited
normal data for MSV and these show considerable variability. Each laboratory should develop its own normal values.
MSV can be difficult to interpret (e.g., in patients with
COPD). The incremental technique may prove to be of
value in the clinical setting. To date, most studies of ventilatory endurance have been undertaken within a research
context.
4. Endurance of external loads applied to the airway: The external load can be resistive (the pressure required depends
on flow), elastic (pressure depends on tidal volume), threshold (finite pressure required to open the valve, which is independent of flow and volume), or an isoflow load (flow
rate held constant). The most widely used technique is that
of threshold loading. Either the maximum sustainable threshold load or the maximum incremental threshold load can
be measured. The incremental threshold loading test, which
uses the same principles as an incremental exercise test, is
the most commonly undertaken, and there are limited normal data available. It is not clear to what extent the test reflects respiratory muscle strength rather than endurance.
5. Repeated maximum inspiratory pressures: In this test the
subjects undertake 18 repeated PI,max maneuvers, each effort lasting 10 seconds with a 5-second rest between contractions. Pressure drops to 87% of PI,max in young normal
subjects over the run. Equipment is simple, and only a manometer and stopwatch are required. Few data from studies in patients are available.
6. Maximal sustainable isoflow: In this test the subject breathes
against a high impedence (air tank) providing a constant
flow (1 L/second). The subject develops maximal pressure
at a TI/Ttot of 0.40. Maximum pressure declines exponentially to a sustainable level of 61%, yielding a PTI of 0.18.
This technique has not yet been tested in large populations.
It has potential as a method of training the inspiratory muscles as well as documenting their endurance.
7. Endurance of the diaphragm: This has been studied in normal
subjects by measuring Pdi and TI/Ttot, which were kept constant by following a pattern of pressure and timing displayed
on an oscilloscope. A PTI of 0.200.30 resulted in task failure
at an earlier time. The technique was developed as a physiologic study designed to measure the use of TTIdi as a parameter to evaluate the development of diaphragm fatigue.
Of the tests of ventilatory endurance available, the most
promising, in a clinical context, appears to be the maximum
incremental ventilation test. To specifically assess the endurance of the inspiratory muscles, relatively independently of
lung and chest wall mechanisms, the incremental threshold
loading test appears to be most useful.
References
1. Bellemare F, Grassino A. Effect of pressure and timing of contraction
on human diaphragm fatigue. J Appl Physiol 1982;53:11901195.
2. Zocchi L, Fitting JW, Majani U, Fracchia C, Rampulla C, Grassino A.
Effect of pressure and timing of contraction on human rib cage muscle
fatigue. Am Rev Respir Dis 1993;147:857864.
3. Schulz L, Nagaraja HN, Rague N, Drake J, Diaz PT. Respiratory muscle
dysfunction associated with human immunodeficiency virus infection.
Am J Respir Crit Care Med 1997;155:10801084.
4. Mancini DM, Henson D, LaManca J, Levine S. Evidence of reduced respiratory muscle endurance in patients with heart failure. J Am Coll
Cardiol 1994;24:972981.
5. Leith DE, Bradley M. Ventilatory muscle strength and endurance training. J Appl Physiol 1976;41:508516.
6. McKenzie DK, Gandevia SC. Strength and endurance of inspiratory, expiratory and limb muscles in asthma. Am Rev Respir Dis 1986;134:9991004.
7. Keens TG, Krastins IRB, Wannamaker EM, Levison H, Crozier DN,
Bryan AC. Ventilatory muscle endurance training in normal subjects
and patients with cystic fibrosis. Am Rev Respir Dis 1977;116:853860.
8. Morrison NJ, Richardson DPT, Dunn L, Pardy RL. Respiratory muscle
performance in normal elderly subjects and patients with COPD.
Chest 1989;95:9094.
9. Weiner I, Azgad Y, Weiner M. Inspiratory muscle training during treatment with corticosteroids in humans. Chest 1995;107:10411044.
10. Homsher E, Kean CJ. Skeletal muscle energetics and metabolism. Annu
Rev Physiol 1978;40:93131.
11. Rall JA. Sense and nonsense about the Fenn effect. Am J Physiol
1982;242:H1H6.
570
12. McCool FD, Leith DE. Mean airway opening pressure as an index of inspiratory muscle task intensity. J Appl Physiol 1986;60:304306.
13. Clanton TL, Ameredes BT, Thomson DB, Julian MW. Sustainable inspiratory pressures over varying flows, volumes, and duty cycles. J
Appl Physiol 1990;69:18751882.
14. Bellemare F, Wight D, Lavigne CM, Grassino A. Effect of tension and
timing of contraction on blood flow of the diaphragm. J Appl Physiol
1986;54:15971606.
15. Collett PW, Perry C, Engel LA. Pressuretime product, flow, and oxygen
cost of resistive breathing in humans. J Appl Physiol 1985;58:12631272.
16. Field S, Sanci S, Grassino A. Respiratory muscle oxygen consumption
estimated by the diaphragm pressuretime index. J Appl Physiol 1984;
57:4451.
17. Begin P, Grassino A. Inspiratory muscle dysfunction and chronic hypercapnia in chronic obstructive disease. Am Rev Respir Dis 1991;143:
905912.
18. Bellemare F, Grassino A. Force reserve of the diaphragm in patients
with chronic obstructive pulmonary disease. J Appl Physiol 1983;55:815.
19. Dodd DS, Kelly S, Collett PW, Engel LA. Pressuretime product, work
rate, and endurance during resistive breathing in humans. J Appl
Physiol 1988;64:13971404.
20. McCool FD, McCann DR, Leith DE, Hoppin FG. Pressureflow effects
on endurance of inspiratory muscles. J Appl Physiol 1986;60:299303.
21. Cala SJ, Edyvean J, Rynn M, Engel LA. O2 cost of breathing: ventilatory
vs. pressure loads. J Appl Physiol 1997;73:17201727.
22. Campbell EJM. The respiratory muscles and the mechanics of breathing.
Chicago, IL: Year Book Publishers; 1958.
23. Banner MJ, Jaeger MJ, Kirby RR. Components of the work of breathing
and implications for monitoring ventilator-dependent patients. Crit
Care Med 1994;22:515523.
24. Roussos C, Campbell EJM. Respiratory muscle energetics. In: Fishman
AP, Macklem PT, Mead J, editors. Handbook of physiology, Section
3: The respiratory system. Vol. III: Mechanics of breathing, Part 2. Bethesda, MD: American Physiological Society; 1986. p. 481509.
25. Roussos C, Zakynthinos S. Respiratory muscle energetics. In: Roussos
C, editor. The thorax. New York: Marcel Dekker; 1997. p. 681749.
26. Tobin MJ, Perez W, Guenther SM, Lodato RF, Dantzker DR. Does rib
cageabdominal paradox signify respiratory muscle fatigue? J Appl
Physiol 1987;63:851860.
27. Goldman MD, Grimby G, Mead J. Mechanical work of breathing derived from rib cage and abdominal VP partitioning. J Appl Physiol
1976;41:752763.
28. De Troyer A, Bastenier-Geens J. Effects of neuromuscular blockade on respiratory mechanics in conscious man. J Appl Physiol 1979;47:11621168.
29. Estenne M, Heilporn A, Delhez L, Yernault JC, De Troyer A. Chest
wall stiffness in patients with chronic respiratory muscle weakness.
Am Rev Respir Dis 1983;128:10021007.
30. Buchler B, Magder S, Katsardis H, Jammes Y, Roussos C. Effects of pleural
pressure and abdominal pressure on diaphragmatic blood flow. J Appl
Physiol 1985;58:691697.
31. Blackie SP, Fairbar MS, McElvaney NG, Wilcox PG, Morrison NJ,
Pardy RL. Normal values and ranges for ventilation and breathing
pattern at maximal exercise. Chest 1991;100:136142.
32. Anholm JD, Johnson RL, Ramanathan M. Changes in cardiac output
during sustained maximal ventilation in humans. J Appl Physiol 1987;
63:181187.
33. Tenney SM, Reese RE. The ability to sustain great breathing efforts.
34. Belman MJ, Mittman C. Ventilatory muscle training improves exercise
capacity in chronic obstructive pulmonary disease patients. Am Rev
Respir Dis 1980;121:273280.
35. Mancini DM, Henson D, La Manca J, Donchez L, Levine S. Benefit of
selective respiratory muscle training on exercise capacity in patients
with chronic congestive heart failure. Circulation 1995;91:320329.
36. American Thoracic Society. Standardization of spirometry: 1994 update.
37. Dillard TA, Piantadosi S, Rajagopal DR. Prediction of ventilation at
maximal exercise in chronic air-flow obstruction. Am Rev Respir Dis
1985;132:230235.
38. Levine S, Weiser P, Gillen J. Evaluation of a ventilatory muscle endurance training program in the rehabilitation of patients with chronic
obstructive pulmonary disease. Am Rev Respir Dis 1986;133:400406.
39. Bai TR, Rabinovitch BJ, Pardy RL. Near-maximal voluntary hyperpnea
40. Hamnegard CH, Wragg SD, Kyroussis D, Mills GH, Polkey MI, Moran
J, Road JD, Bake B, Green M, Moxham J. Diaphragm fatigue following maximal ventilation in man. Eur Respir J 1996;9:241247.
VOL 166
2002
41. Mador JM, Rodis A, Diaz J. Diaphragmatic fatigue following voluntary

hyperpnea. Am J Respir Crit Care Med 1996;154:6367.
42. Polkey MI, Kyroussis D, Hamnegard CH, Mills GH, Hughes PD, Green
M, Moxham J. Diaphragm performance during maximal voluntary
ventilation in chronic obstructive pulmonary disease. Am J Respir Crit
Care Med 1997;155:642648.
43. Fitting JW, Bradley TD, Easton PA, Lincoln MJ, Goldman MD,
Grassino A. Dissociation between diaphragmatic and rib cage muscle
fatigue. J Appl Physiol 1988;64:959965.
44. Hershenson MB, Kikuchi Y, Tzelepis GE, McCool D. Preferential fatigue of the rib cage muscles during inspiratory resistive loaded ventilation. J Appl Physiol 1997;66:750754.
45. Roussos C, Gross D, Macklem PT. Fatigue of inspiratory muscles and
their synergic behavior. J Appl Physiol 1979;46:897904.
46. Nickerson BG, Keens TG. Measuring ventilatory muscle endurance in
humans as sustainable inspiratory pressure. J Appl Physiol 1982;52:
768772.
47. Clanton TL, Dixon GF, Drake J, Gadek JE. Effects of swim training on
lung volumes and inspiratory muscle conditioning. J Appl Physiol 1987;
62:3946.
48. Clanton TL, Dixon G, Drake J, Gadek JE. Inspiratory muscle conditioning using a threshold loading device. Chest 1985;87:6266.
49. Eastwood PR, Hillman DR. A threshold loading device for testing of inspiratory muscle performance. Eur Respir J 1995;8:463466.
50. Martyn JB, Moreno RH, Pare PD, Pardy RL. Measurement of inspiratory muscle performance with incremental threshold loading. Am Rev
Respir Dis 1987;135:919923.
51. Belman MJ, Scott GT, Lewis MI. Resistive breathing training in patients
with chronic obstructive pulmonary disease. Chest 1986;90:662669.
52. Clanton TL, Dixon GF, Drake J, Gadek JE. Effects of breathing pattern
on inspiratory muscle endurance in humans. J Appl Physiol 1985;59:
18341841.
53. McElvaney G, Fairbarn MS, Wilcox PG, Pardy RL. Comparison of twominute incremental threshold loading and maximal loading as measures
of respiratory muscle endurance. Am Rev Respir Dis 1989;96:557563.
54. Tolep K, Kelsen SG. Effect of aging on respiratory skeletal muscles. Clin
Chest Med 1993;14:363378.
55. Weiner P, Azgad Y, Ganam R. Inspiratory muscle training combined
with general exercise reconditioning in patients with COPD. Chest
1992;102:13511356.
56. Morrison NJ, Fairbarn MS, Pardy RL. The effect of breathing frequency
on inspiratory muscle endurance during incremental threshold loading. Chest 1989;96:8588.
57. Eastwood PR, Hillman DR, Finucane KE. Ventilatory responses to inspiratory threshold loading and role of muscle fatigue in task failure. J
Appl Physiol 1994;76:185195.
58. McKenzie DK, Plassman BL, Gandevia SC. Maximal activation of the
human diaphragm but not inspiratory intercostal muscles during static
inspiratory efforts. Neurosci Lett 1988;89:6368.
59. Ameredes BT, Clanton TL. Hyperoxia and moderate hypoxia fail to affect
inspiratory muscle fatigue in humans. J Appl Physiol 1989;66:894900.
60. Juan G, Claverley P, Talamo C, Schnader J, Roussos C. Effect of carbon
dioxide on diaphragmatic function in humans. N Engl J Med 1984;310:
874879.
61. Ameredes BT, Clanton TL. Accelerated decay of inspiratory pressure
during hypercapnic endurance trials in humans. J Appl Physiol 1988;
65:728735.
62. McKenzie DK, Gandevia SC. Influence of muscle length on human inspiratory and limb muscle endurance. Respir Physiol 1987;67:171182.
63. Gandevia SC, McKenzie DK, Neering IR. Endurance properties of respiratory and limb muscles. Respir Physiol 1983;53:4761.
64. Clanton TL, Ameredes BT. Fatigue of the inspiratory muscle pump in
humans: an isoflow approach. J Appl Physiol 1988;64:16931699.
65. Clanton TL, Hartman E, Julian MW. Preservation of sustainable inspiratory muscle pressure at increased end-expiratory lung volume. Am
Rev Respir Dis 1992;147:385391.
66. Mador MJ, Rodis A, Magalang UJ, Ameen K. Comparison of cervical
magnetic and transcutaneous phrenic nerve stimulation before and after threshold loading. Am J Respir Crit Care Med 1996;154:448453.
67. Bellemare F, Grassino A. Evaluation of human diaphragm fatigue. J
Appl Physiol 1982;53:11961206.
68. Vassilakopoulos T, Zakynthinos S, Roussos C. The tensiontime index
and the frequency/tidal volume ratio are the major pathophysiological
determinants of weaning failure and success. Am J Respir Crit Care
Med 1998;158:378385.
69. Belman MJ, Gaesser GA. Ventilatory muscle training in the elderly. J
Appl Physiol 1988;64:899905.
5. Assessment of Respiratory Muscle Fatigue

The purpose of this Section is to outline a potential diagnostic
strategy to assess the development of respiratory muscle fatigue
in humans. At the outset, it is important to note that there is evidence suggesting that human respiratory muscle fatigue may
develop in pathophysiological states associated with the development of respiratory failure (i.e., respiratory loading induced
by lung disease) (1). It is also important to recognize that fatigue is defined as a loss of the capability to generate skeletal
muscle force and/or velocity that is accompanied by recovery
during rest (2). As a result, a single measurement of force is
inadequate to detect fatigue: rather, muscle force generating
or shortening capability must be demonstrated to fall during
serial measurements over time. It could further be argued that
a demonstration that force subsequently rises if muscle contraction is stopped and a rest period is provided would be necessary to fully satisfy the definition of fatigue and to exclude
the possibility that a given fall in force did not represent muscle injury (the latter condition, by definition, does not improve
with short periods of rest). It follows that muscle fatigue can
therefore be distinguished from muscle weakness (i.e., a reduction in the level of force generation at a given point in time)
and muscle injury (i.e., a slowly reversible or irreversible decrement in muscle contractility).
This Section is divided into a brief discussion of (1) the
present theories regarding the genesis of muscle fatigue and
the different types of muscle fatigue, and (2) a review of the
various tests available with the potential to detect the development of fatigue in normal subjects and patients.
TYPES OF FATIGUE
On an operational level, it has proven convenient to classify
fatigue into different types, with these different forms of fatigue representing different biophysical mechanisms of fatigue
development and with each type having different physiological characteristics (3). Several such classification schemes are
possible, but a widely employed convention is to classify fatigue as either (1) central fatigue, (2) peripheral high-frequency fatigue, or (3) peripheral low-frequency fatigue. We
review each of these types of fatigue in turn.
Central Fatigue
Central fatigue refers to a condition in which muscle force

generation during sustained or repetitive contraction becomes
limited owing to a decline in motoneuronal output. Central
fatigue is judged to be present when a truly maximum voluntary effort produces less force than one generated by direct
electrical stimulation.
A number of experiments have suggested that a form of
central diaphragmatic fatigue may develop during respiratory loading (311). A study by Bellemare and BiglandRitchie (5) provided evidence that such a phenomenon can
occur during the application of external resistive loads to normal human subjects. This study measured transdiaphragmatic
pressure generation over time before, during, and after inspiratory resistive loading, and employed superimposed electrical phrenic stimulation at various times during the experiment to determine whether subjects were capable of fully
activating the diaphragm. This approach makes use of the
observation that it is possible for well-motivated individuals to
fully
activate rested skeletal muscle during volitional contracDOI: 10.1164/rccm.AT0502
Internet
address:
www.atsjournals.org
tions when
making
a maximal effort (i.e., superimposed elec-
trical stimulation of muscle during such maximal maneuvers

does not result in an increase in force generation above that
achieved volitionally) (3) (see TWITCH OCCLUSION in Section 2
of this Statement). Although achievement of such maximal
activation is difficult, and usually cannot be achieved with every attempted contraction even in motivated individuals, one
study found that all research subjects could achieve at least
one maximal contraction of a limb muscle (6). As a result, evidence that maximal activation of a given muscle during a maximal volitional effort can never be achieved after a period of
exercise (i.e., the superimposed electrical stimulation can always evoke an increase in force generation) constitutes evidence of central fatigue.
At the start of the study by Bellemare and Bigland-Ritchie,
no superimposed force could be detected during the imposition of electrical stimuli on maximum volitional efforts, indicating that these subjects were capable of maximally activating the diaphragm before respiratory loading. During the
course of loading, however, the extent to which the diaphragm
could be activated decreased progressively. Evidence of the
development of central diaphragmatic fatigue during repeated
maximal and submaximal diaphragmatic contractions has also
been reported in a study by McKenzie and coworkers (7).
Other work has suggested that central fatigue may be the
result of a decrease in central respiratory motor outflow in response to opioid elaboration in the central nervous system,
with the latter generated, in turn, as a consequence of the
stress of loaded breathing (4, 8, 9). In support of this concept,
Santiago and coworkers (8) have shown that naloxone restores the load compensatory reflex in patients with chronic
obstructive pulmonary disease in whom it is initially absent.
Subsequently, this group demonstrated that resistive loading
in unanesthetized goats resulted in a progressive reduction in
tidal volume, which was partially reversed by administration
of naloxone (4).
In keeping with the results of these animal studies, an experiment involving patients with asthma found that naloxone
pretreatment alters the response to methacholine challenge.
In these individuals, in whom methacholine induced severe reductions in FEV1, naloxone pretreatment resulted in an increased breathing frequency, occlusion pressure, and mean inspiratory flow rate when compared with saline pretreatment
(10). It has been postulated that similar central limitations of
respiratory motor outflow may occur in patients with diseases
that chronically load the respiratory system, contributing to
the development of chronic hypercapnia.
High-Frequency Peripheral Fatigue
Central fatigue is a failure to generate force as a result of a reduction in motor output from the central nervous system. Peripheral fatigue refers to failure at the neuromuscular junction
or distal to this structure and is judged to be present when
muscle force output or velocity falls in response to direct electrical stimulation. Peripheral fatigue can result because of alterations in the neuromuscular junction, changes in propagation of the action potentials along the sarcolemmal membrane
or into the t-tubules, changes in excitationcontraction coupling, or because of other alterations within the muscle cell
(e.g., alterations in metabolism, changes in contractile proteins). Peripheral fatigue can be further classified into highfrequency and low-frequency fatigue on the basis of the shape
of the postfatigue muscle forcefrequency relationship. If fa-
572
VOL 166
2002
tigue results in depression of the forces generated by a muscle

in response to high-frequency electrical stimulation (e.g., in
humans, 50100 Hz) then high-frequency fatigue is said to be
present, whereas a reduction in force generation in response
to low-frequency stimuli (i.e., 120 Hz) is taken as an indication of low-frequency fatigue. Studies have suggested that loss
of force at low frequencies represents an impairment of muscle excitationcontraction coupling (i.e., a reduction in contractile protein activation in response to a given nonimpaired
sarcolemmal action potential) (12). Reduction in high-frequency force generation is thought to indicate either an alteration in neuromuscular junction transmission, a reduction in
sarcolemmal membrane excitability, or a reduction in action
potential propagation into the t-tubular system (13, 14). Lowfrequency fatigue can occur in isolation, but high-frequency
fatigue is invariably associated with some alterations in muscle
force generation at lower frequencies.
High-frequency fatigue has been demonstrated in the diaphragms of normal humans after a trial of high-intensity inspiratory resistive loading (this was demonstrated using highfrequency electrical phrenic stimulation) (15). In this study
high-frequency fatigue resolves extremely quickly after cessation of strenuous muscle contractions (i.e., after removal of
the inspiratory resistive load; see Figure 1).
ble of detecting another. Moreover, the necessity to make serial measurements of an index of muscle force generation over
time to detect fatigue is a particularly difficult endeavor for
the respiratory system, because a large number of variables
(i.e., lung volume, thoracoabdominal configuration, muscle interaction) can vary over time. All these factors can influence
the relationship between muscle force and pressure generation.
As an example, consider the utility of measuring maximum
inspiratory pressure (PI,max) serially to detect respiratory
muscle fatigue. This parameter is highly effort dependent, and
time-dependent reductions could represent lack of motivation,
central fatigue, peripheral high-frequency fatigue, or simply an
alteration in lung volume and a resultant mechanical change in
the transduction of muscle force into pressure. In addition, failure of the PI,max to change does not exclude the development
of fatigue, because this test would not be suitable to detect lowfrequency fatigue. As a result, one must keep in mind the potential limitations of a given test for the detection of muscle fatigue. Most tests are suitable for detecting the presence of only
one component of muscle fatigue, and complete characterization of fatigue requires a complex series of assessments.
Low-Frequency Peripheral Fatigue
Rationale. Rapid shallow breathing, characterized by high breathing frequency and low tidal volume, commonly develops in progressive respiratory failure or in unsuccessful attempts to wean
from mechanical ventilation. These conditions are associated
with an increased ventilatory load and/or a reduced respiratory muscle capacity and may therefore potentially lead to respiratory muscle fatigue (see PREDICTION OF WEANING in Section 10 of this Statement).
Methodology and equipment. Breathing frequency can be
easily counted at the bedside and should be included in standard monitoring. Tidal volume measurements of intubated patients are commonly accomplished with the flow sensors built
into modern ventilatory equipment and can be displayed on a
breath-by-breath basis by these machines. Volume measurements can also be made with Wright respirometers and other
spirometric devices via a mouthpiece in nonintubated patients,
albeit mouthpiece placement can artifactually alter tidal volumes and respiratory patterns. To avoid such artifacts, it is
possible to noninvasively monitor tidal volume by respiratory
inductance plethysmography. Use of this and similar methods
is described in DEVICES USED TO MONITOR BREATHING: PNEUMOGRAPH, MAGNETOMETER, AND RESPIRATORY INDUCTIVE PLETHYSMOGRAPH in Section 6 of this Statement.
Advantages. Monitoring changes in breathing frequency
and tidal volume is simple and noninvasive.
Disadvantages. The relationship between fatigue and breathing pattern is complex. Moreover, rapid shallow breathing is
most likely a reflex response to an increase in the respiratory
workload (24) and not the consequence of respiratory muscle
fatigue per se (25). Thus, although rapid shallow breathing may
accompany respiratory muscle fatigue (2), it cannot be considered a specific marker of fatigue.
Applications. Monitoring breathing frequency and tidal volume represents a part of the routine respiratory surveillance of
patients, but these parameters should not be used as specific indicators of the development of respiratory muscle fatigue (see
BREATHING PATTERN in Section 10 of this Statement).
In the presence of pure low-frequency fatigue, force generation in response to high-frequency stimulation is unimpaired,
indicating that the contractile proteins are capable of generating maximal force provided that sufficient calcium is released
by the sarcoplasmic reticulum (SR). As a result, impaired
force generation at submaximal frequencies of stimulation
may represent either a reduced level of calcium availability
due to alterations in SR function or a reduction in the calcium
sensitivity of the myofilaments at submaximal calcium concentrations. Both changes have been demonstrated experimentally (16, 17). Reduced myofilament calcium sensitivity can be
produced experimentally by increasing hydrogen and phosphate ions (18). The explanation for impaired calcium release
by the SR during contractions is less well understood, and a
number of theories have been proposed to account for this
phenomenon (16, 1921).
Low-frequency fatigue has been demonstrated in the diaphragm and sternocleidomastoid muscles of normal subjects
breathing against high resistive loads (17, 22). Low-frequency
fatigue has also been shown to develop in the diaphragm of
normal subjects asked to sustain maximum voluntary ventilation for 2 minutes (23).
Implications of Different Types of Fatigue for Diagnosis
Although it is convenient to discuss the characteristics of central, peripheral high-frequency, and peripheral low-frequency
fatigue separately, it is likely that these various phenomena do
not occur in isolation during muscle activation. All three processes may be operating simultaneously when the respiratory
muscles confront an excessive workload, with the relative importance of each depending on the duration of respiratory
loading and other physiological variables (i.e., arterial pressure, arterial blood gas concentrations, nutritional state). Whereas
all three processes may participate in the acute response to
loading, both central and high-frequency fatigue resolve rapidly once fatiguing levels of muscle contraction cease, and only
low-frequency fatigue is likely to persist over minutes to hours.
Because muscle fatigue is a complex phenomenon, a test
that is well suited to detect one form of fatigue may be incapa-
TESTS OF RESPIRATORY MUSCLE FATIGUE

Breathing Pattern: Tidal Volume and Breathing Frequency
Thoracoabdominal Motion
Rationale. The analysis of breathing movements gives some insight into the level of recruitment and function of the respiratory
573
Figure 1. Pressurefrequency relationships for the human diaphragm

at various time points after a loaded breathing trial (i.e., a breathing
trial in which an external resistive load was applied to increase the
work done by the diaphragm and simulate the effects of lung disease). Diaphragm pressure generation was measured by determining the transdiaphragmatic pressure gradient (Pdi) generated in response to electrical stimulation of the phrenic nerves; Pdi is shown
on the y axis and phrenic stimulation frequency is shown on the x
axis. The time at which each set of curves was obtained is shown in
the upper left-hand corner of each panel. For reference, each panel
contains both a postload pressurefrequency curve obtained at the
designated time (dashed line), and a preload pressurefrequency
curve (solid line). At 24 minutes postloading, there was a large decrement in pressure generation in response to phrenic stimulation of
20100 Hz (top left panel), indicating diaphragm fatigue. Over time,
the pressures generated in response to high-frequency stimulation
increased so that at 2530 minutes postloading, pressure generation
in response to 50- and 100-Hz stimulation had returned to preloading levels (bottom right panel). In contrast, however, response to 20Hz stimulation remained depressed 2530 minutes after loading. In effect, both low- and high-frequency peripheral diaphragm fatigue
was present 24 minutes after loading, with rapid resolution of the
high-frequency component of fatigue and persistence of the lowfrequency component at 2530 minutes. Reprinted by permission
from Reference 15.
muscles (in particular of the diaphragm, the rib cage inspiratory

muscles, and the abdominal muscles). Two unusual patterns of
muscle recruitment may be observed in healthy subjects subjected to fatiguing inspiratory loads (26). The first is an increased
variability in compartmental contribution to tidal volume, with
breaths characterized by clear rib cage predominance alternating with other breaths in which abdominal motion predominates. This pattern reflects alternatively predominant recruitment of the inspiratory rib cage muscles and of the diaphragm.
Because fatigue may develop separately in the diaphragm and in
the inspiratory rib cage muscles (27), such alternation may represent a way to postpone respiratory muscle failure. The second
pattern is frank paradoxical movement of one compartment,
generally the abdomen, that is, an inward movement of the abdominal wall during inspiration. Abdominal paradox indicates
weak, absent, or inefficient contraction of the diaphragm. These
two patterns may also be observed in patients showing signs of
diaphragmatic fatigue during weaning trials from mechanical
ventilation (2) (see Figure 5 in Section 6 of this Statement).
Methodology. Most anomalies of thoracoabdominal motion can be detected by visual inspection by a trained observer. This assessment is facilitated by placing the patient in a
recumbent position and conducting a visual inspection for several minutes. Quantitative measurements of rib cageabdominal motion can also be performed (see ESTIMATION OF VENTILATION BASED ON CHEST WALL MOTION: KONNOMEAD DIAGRAM
in Section 6 of this Statement).
Advantages. Visual inspection provides a simple bedside
means of detecting alterations in respiratory muscle use.
Disadvantages. The abnormal patterns of thoracoabdominal motion described above are not specific for respiratory
muscle fatigue. Indeed, respiratory alternans and abdominal
paradox can appear immediately after the institution of loaded
breathing and these abnormalities do not appear to become
accentuated with the development of fatigue. Furthermore,
these patterns can also occur, albeit to a lesser degree, during
the application of low, nonfatiguing respiratory loads (28).
Thus, abnormal thoracoabdominal motion should be viewed
as reflecting an increased ventilatory load, which in itself may
or may not induce respiratory muscle fatigue.
Applications. Analysis of thoracoabdominal motion is most

useful to detect either specific forms of respiratory muscle
dysfunction (e.g., diaphragmatic paresis) and/or an increase in
the ventilatory load. This assessment is, therefore, of some
routine clinical use, but lacks specificity for detecting respiratory muscle fatigue (see BREATHING PATTERN in Section 10 and
DEVICES USED TO MONITOR BREATHING: PNEUMOGRAPH, MAGNETOMETER, AND RESPIRATORY INDUCTIVE PLETHYSMOGRAPH in
PressureTime Index of Inspiratory Muscles
Rationale. In several investigations, skeletal muscle fatigue

was found to occur when a muscle generated more than 15%
of its maximal force during sustained contraction. This work
has led to the concept that a fatigue threshold exists, with fatigue occurring only when the level of pressure-time generated exceeds this threshold level. Additional work on this concept has shown that the fatigue threshold is higher during
intermittent contractions and depends on the relative duration
of contraction and relaxation. The same holds true for the inspiratory muscles of subjects submitted to external inspiratory
loads. The pressuretime index of the diaphragm is defined as
PTdi = ( Pdi Pdi,max ) ( T I Ttot )
where Pdi is the mean transdiaphragmatic pressure generated
per breath, Pdi,max is maximal transdiaphragmatic pressure,
TI is inspiratory time, and Ttot is total breath time. When
breathing is accomplished predominantly with the diaphragm,
the critical PTdi is 0.150.18. Below this threshold, breathing
can be sustained for more than 1 hour without evidence of fatigue. Above this threshold, task failure occurs after a time
limit that is inversely related to PTdi (29) (see Figure 2). In
most situations in which inspiratory loads are applied, the
spontaneous breathing pattern is characterized by predominant recruitment of inspiratory rib cage muscles other than
the diaphragm, resulting in augmented rib cage expansion and
abdominal paradox. Under these circumstances, the pressure
time index of the inspiratory rib cage muscles is defined as
PTrc = ( Ppl Ppl,max ) ( T I Ttot )
574
where Ppl is mean pleural pressure generated per breath and

Ppl,max is maximal pleural pressure (equivalent to MIP).
With this breathing pattern, the critical PTrc is 0.30. Above
this threshold, task failure occurs after a time limit that is inversely related to PTrc (30).
Methodology and equipment. The measurement of esophageal pressure is required to compute PTrc, and the measurement of both esophageal and gastric pressures is required to
compute PTdi. This is most commonly performed with balloon-catheter systems, as described in PRESSURE MEASUREMENTS in Section 2 and in Figures 8 and 9 in Section 4 of this
Statement.
Advantages. In principle, PTdi and PTrc characterize the
operational conditions of the inspiratory muscles with respect
to their fatigue threshold. These indices may allow the assessment of the risk of fatigue before actual task failure occurs.
Disadvantages. The critical values of PTdi and PTrc were
established in healthy subjects breathing against external
loads. The critical thresholds may be different in clinical circumstances, in which a number of pathological factors (e.g.,
levels of tissue perfusion, presence of hypoxemia) may influence muscle performance. Second, pressuretime index assessment is dependent on accurate measurement of maximal
muscle pressure generating capacity (i.e., the Pdi,max of the
diaphragm, Ppl,max for the inspiratory rib cage musculature),
which is often difficult in patients. Third, shortening velocity
of muscle fibers strongly influences muscle energetics and the
metabolic consequences of contraction. As a result, the critical
PTdi and PTrc that can be tolerated are also a function of inspiratory flow patterns, with lower values for these parameters
at high inspiratory flows. Fourth, some clinical conditions
(malnutrition, steroid myopathy) change muscle fiber populations, altering the relationship between muscle strength and
fatigability. Conditions that result in a shift to a greater concentration of slow fibers in muscle may well result in a better
tolerance of a given absolute level of the pressuretime index
and an increase in the critical pressuretime index for the diaphragm and rib cage muscles. The critical pressuretime index
in most patients remains to be measured.
Applications. The pressuretime index should be considered a conceptual framework within which to gauge the level
of muscle function rather than an instrument for the clinical
diagnosis of fatigue. Fatigue thresholds have been reported in
patients with chronic obstructive pulmonary disease (31, 32);
however, it remains largely untested in other pathologies.
Volitional Maximal Pressures
Rationale. Skeletal muscle fatigue has been defined as a loss of

capacity to develop force in response to a load that is reversible by rest (12). In accordance with this definition, respiratory
muscle fatigue can, potentially, be documented by measuring
a decrease in volitional maximal respiratory pressures, with
demonstration of recovery with rest. As a consequence, to detect fatigue of the inspiratory muscles, one could measure either maximal static inspiratory pressure, maximal transdiaphragmatic pressure, or maximal sniff pressure.
Maximal inspiratory pressure. Methodology and equipment.
PI,max is measured at the mouth as described in VOLITIONAL
TESTS OF RESPIRATORY MUSCLE STRENGTH in Section 2 of this
Statement.
Advantages. The measurement of PI,max is noninvasive.
Fatigue of inspiratory muscles as a whole has been documented by a transient fall of PI,max after breathing against external loads (12), maximal voluntary hyperpnea (33), marathon running (34, 35), or labor (36).
VOL 166
2002
Figure 2. Data obtained from human subjects generating a variety of

levels of transdiaphragmatic pressure (Pdi/Pdi,max) with various inspiratory durations (TI/Ttot) during breathing trials. Each point represents the results from a single breathing trial that was performed at a
targeted Pdi/Pdi,max (x axis) and with a pattern of breathing that resulted in a predetermined inspiratory duration (TI/Ttot isopleths of 1,
0.65, 0.5, and 0.25). Subjects continued breathing trials until they
were unable to sustain the targeted pressure level; the duration each
trial could be sustained was defined as Tlim (shown on the y axis). This
graph shows that the time breathing could be sustained (Tlim) was a
function of both the pressure developed (Pdi/Pdi,max) and the inspiratory duration (TI/Ttot). For a given TI/Ttot isopleth, Tlim was shorter
when pressure generation was higher and, for a given pressure generation level, Tlim was shorter when inspiratory duration was increased.
Disadvantages. The major limitation of PI,max as a test of

fatigue is a lack of specificity, which is due to its total dependence on the subjects maximal volitional effort. Although
PI,max is reliable in highly motivated subjects, a maximal volitional effort cannot be obtained with certainty in patients. Another drawback of PI,max is a potential lack of sensitivity for
fatigue. Because a maximal static effort is associated with a
high neuronal firing frequency, it reflects mainly high-frequency fatigue and may be a poor indicator of long-lasting
low-frequency fatigue (see MAXIMAL STATIC INSPIRATORY AND
EXPIRATORY PRESSURE in Section 2 of this Statement).
Applications. Measurement of PI,max can be used to detect
inspiratory muscle fatigue in motivated volunteers, but has limited use in patients for this purpose because of difficulties in ensuring maximality of effort. Twitch interpolation techniques provide a potential means of solving this latter problem (see TWITCH
OCCLUSION in Section 2 and see Section 1 of this Statement).
Maximal transdiaphragmatic pressure. Methodology. Pdi,max,
measured with balloon-catheter systems, is described in TECHNIQUES FOR PRESSURE MEASUREMENTS in Section 2 of this Statement).
Advantages. This test measures specifically the strength of
the diaphragm. Diaphragmatic fatigue has been documented
by a transient fall in Pdi,max after breathing with diaphragm
emphasis against external loads (6), voluntary hyperpnea (33),
or high-intensity exercise (37).
Disadvantages. The measurement of Pdi,max is invasive and,
moreover, shares the same type of limitations as PI,max for detecting fatigue. Because the physical maneuver that patients must
carry out for reliable measurement of Pdi,max is even more complex than that required to measure the PI,max, it cannot be recommended for assessing diaphragmatic fatigue in clinical settings.
Applications. Pdi,max can be used to detect inspiratory

muscle fatigue in motivated volunteers, but probably should
not be used for this purpose in patients.
Maximal sniff pressures. Methodology. Diaphragmatic strength
can be assessed by maximal sniff Pdi (38), and global inspiratory muscle strength by maximal sniff esophageal pressure
(Pes) (39) or maximal sniff nasal inspiratory pressure (SNIP)
(40, 41), as described in SNIFF TESTS in Section 2 of this Statement.
Advantages. The sniff is a volitional maneuver that is easily
performed by almost all subjects and patients. The SNIP is
noninvasive and often yields higher pressures than the PI,max
(39).
Disadvantages. The potential usefulness of measuring a fall
of maximal sniff pressures to detect inspiratory muscle fatigue
in patients remains to be established.
Applications. Although maximal sniff pressures are of established value for measuring inspiratory muscle strength, it is
a test still under development for documenting fatigue.
Relaxation Rate
Rationale. On cessation of contraction, skeletal muscles relax

at a rate determined by their relative proportion of fast and
slow fibers. When muscles fatigue, their relaxation rate declines as a result of a slower uptake of calcium previously released from the sarcoplasmic reticulum. During various types
of intermittent contractions, the rate of decay of Pes and of
Pdi reflects the relaxation rate of inspiratory muscles and of
the diaphragm, respectively. When fatigue is induced by
breathing against external loads, the inspiratory muscle relaxation rate falls early and then stabilizes, following a time
course similar to that of the change in electromyogram
(EMG) power spectrum. Thus, relaxation rates typically decline before the occurrence of muscle failure at about the
same rate as the center frequency of EMG does. On cessation
of loading, the relaxation rate recovers quickly and reaches
baseline values within 5 to 10 minutes (42, 43).
Methodology. The relaxation rate of Pes or Pdi can be
measured during intermittent contractions against loads (42),
during sniffs with airway occlusion (43) or without airway occlusion (44), and during phrenic nerve stimulation (43). The
most useful and simple maneuver is the unoccluded sniff,
which is easy to perform for most subjects and provides large
and consistent changes in relaxation rate after fatigue (45).
Standard balloon-catheter systems are used to measure Pes
and gastric pressure (Pga), from which Pdi is obtained. The
maximal relaxation rate (MRR) of Pes or Pdi is calculated as
the first derivative of pressure with respect to time (dP/dt)
over the first half of the relaxation curve. This is obtained by
drawing a tangent to the steepest portion of the pressure
curve. Because the MRR increases with the amplitude of the
pressure swing, it is usual to normalize the MRR and to express it as a percentage of the pressure fall in 10 milliseconds
(42, 45). When the natural logarithm of pressure is plotted as a
function of time, a straight line appears over the lower 60
70% portion, indicating a monoexponential decay. The reciprocal of the slope of this line represents the time constant ()
of this exponential decay, which may be used as another measure of muscle relaxation, usually expressed in milliseconds
(42, 45) (see Figure 3). Thus, a slower muscle relaxation is documented by a decline in the MRR and by an increase in . Inspiratory muscle relaxation rate can be assessed in a less invasive manner by measuring the MRR of nasopharyngeal or
mouth pressure during sniffs with balloons positioned in these
locations (44). An entirely noninvasive measure of inspiratory
muscle MRR can be obtained by using SNIP (46). These less
575
invasive techniques have been validated only in normal subjects. Transmission of brief pressure swings from the alveoli to
the upper airways is likely to be dampened in patients with abnormal lung mechanics.
Equipment. The pressure measurement system required
for sniffs is described in Section 2 of this Statement. Analysis
of the MRR is best done with a computer program (46).
Advantages. The measurement of inspiratory muscle relaxation rate is relatively simple and requires minimal cooperation from subjects. The sniff maneuver is easily performed by
most subjects and patients and does not need to be perfectly
maximal, provided that the MRR is expressed as percentage
pressure fall per 10 milliseconds. Sniffs should, however, be
performed as near to the maximal as possible, because the
MRR is effort dependent below 60% of maximal pressure
(47). The muscle relaxation rate slows at an early stage during
fatiguing contractions and may therefore indicate that inspiratory muscle fatigue is incipient.
Disadvantages. The relationship between changes in relaxation rate and force loss during fatigue is not understood. For
instance, the degree of force loss does not correlate with the
changes in relaxation rate and therefore cannot be inferred
from this parameter (45, 47). The rapid recovery of this index
with rest also poses practical problems of measurement in
clinical settings. The range of normal values for the MRR and
is wide, with overlap between fresh and fatigued states. Serial measurements are thus required to detect the onset of inspiratory muscle fatigue in an individual (45). Finally, some
clinical conditions (e.g., asthma) have been reported to elicit
activation of inspiratory muscles during expiration (postinspiratory inspiratory muscle activity). Under such circumstances,
persistent activation of some muscles or portions thereof would
Figure 3. The time constant of respiratory muscle relaxation, , as a

function of breath number during a series of breathing trials against
respiratory loads. Each curve represents the results from a single
breathing trial, with individual points on each curve representing relaxation time constants calculated on single breaths during the trial.
Each trial is annotated with the pressuretime index (TTdi, equivalent
to the pressuretime index using the nomenclature in the present document) achieved during the trial. During a trial with a high pressure
time index of 0.4 (uppermost curve), relaxation slowed dramatically,
with a large increase in the relaxation time constant, , within a few
breaths. In contrast, with less intense respiratory loading, that is, lower
pressuretime indices, there was less slowing of relaxation and smaller
increases in the relaxation time constant. At the lowest pressuretime
index (0.15, lowest curve), there was essentially no slowing of relaxation and no increase in the time constant during the breathing trial.
576
be expected to alter measured relaxation rates, distorting the

relationship of alterations in the MRR to cellular events and
to the development of muscle fatigue.
Applications. Measurement of the relaxation rate can be
used with confidence as an early sign of fatigue in subjects
subjected to high external inspiratory loads (4245) or to highlevel hyperpnea (46). The inspiratory muscle relaxation rate
can also be used to detect fatiguing contractions during exercise in patients with chronic obstructive pulmonary disease
(48) or during weaning trials from mechanical ventilation (49).
Because interpretation of changes in relaxation rate is not
straightforward, this test can be considered useful only for clinical research.
Electromyography
Time domain analysis. Rationale. For the respiratory muscles, as for any other skeletal muscles, a nearly linear relationship may be found between the pressure and the electrical activity they generate. The slope is related to force and the
length of the diaphragm. As a result, for a given muscle length,
a decrease in the ratio of respiratory muscle pressure to the integrated electromyographic activity of the muscle generating that
pressure should, in theory, indicate a decrease in muscle contractility and the development of fatigue. Furthermore, it has been
suggested that a decrease in this ratio indicates an alteration in
excitationcontraction coupling (50). This point is also discussed
in INFERRING DIAPHRAGM ACTIVATION AND ELECTROMECHANICAL
EFFECTIVENESS FROM EMG in Section 6 of this Statement.
Methodology. The methodology required for electromyographic assessment is reviewed in detail in EMG EQUIPMENT
AND DATA ANALYSIS in Section 3 of this Statement.
Advantages. In theory, this is a useful approach for separating changes in pressure-generating capacity caused by neural or neuromuscular transmission factors from changes caused
by peripheral muscular factors (51). One potential major advantage of this test is the possibility to detect fatigue during
spontaneous breathing (52), because no special efforts are required by patients.
Disadvantages. For this index to be valid, other factors affecting respiratory muscle contractility, such as muscle length,
chest wall configuration, or lung volume, should be controlled or
kept constant (53). The applicability of this particular method
to the respiratory system is limited by the difficulty of recording the activity of all the muscles involved in normal or augmented breathing that contribute to the measured pressure.
Their relative contribution to the generated pressure is known
to change during fatigue development (6, 28), and a reliable recording of the activity of a selective respiratory muscle group
is regarded as difficult by some (54). Section 3 of this Statement offers a more optimistic view. In practice, the diaphragm,
the neck accessory muscles, and the abdominal muscles are
most amenable to this form of testing because their electrical activity can be more easily recorded without interference from
other muscles and their force production (sternomastoid) or
pressure output (diaphragm and abdominals) can also be recorded in relative isolation.
When interpreting results, one must also recognize that the
relationships between integrated EMG activity of the respiratory muscles and the pressure they generate may not be perfectly linear (49).
If special precautions are not taken, EMG signals (particularly those recorded from the diaphragm with an esophageal
electrode) can be subject to artifactual changes caused by variations in lung volume or chest wall configuration (55). Luckily,
reports provide techniques that exclude many of the artifacts
associated with esophageal diaphragmatic recording. Specifically,
VOL 166
2002
work by Sinderby, Grassino, and others provides a means of

recording and analyzing electromyographic signals so as to exclude electrocardiogram (ECG), electrode motion, noise, and
esophageal peristalsis artifacts (5659). This latter work has also
shown that it is possible, by using a multielectrode array, to reliably measure the diaphragm EMG amplitude and power spectrum in such a way that these variables are not affected by chest
wall configuration and/or diaphragm length (56). Note, however, that even if EMG activity can be accurately recorded respiratory muscle pressures must also be reliably assessed for
the pressure-to-integrated EMG ratio to be meaningful.
Applications. The theoretical value of time domain electromyographic assessment in patients is that it can provide a means of
determining whether observed reductions in respiratory muscle pressure-generating capacity are due to alterations in action
potential transmission or to intrinsic alterations in peripheral
muscle function (i.e., alterations in excitationcontraction coupling or contractile protein myofilament function). Although
this technique is principally of value for experimental applications at the present time, ongoing efforts are being made to
improve the reliability of this form of testing, and this type of
measurement may assume a broader clinical role in the future.
Frequency domain analysis. Rationale. Frequency domain
analysis of EMG signals from the respiratory muscles has
been proposed as a test to detect the occurrence of respiratory
muscle fatigue in humans (60), because the power spectrum of
skin surface-recorded EMG signals typically shifts to lower
frequencies during fatiguing contractions (see Figure 4).
Several indices of the power spectrum have been used for
this purpose, including an assessment of the center or centroid frequency of the power spectrum and a power ratio
of a high-frequency band over a low-frequency one. Both of
these indices appear to decrease with fatigue and increase
with recovery. With appropriate instrumentation, these analyses can be obtained on line in spontaneously breathing subjects or patients. Shifts in the EMG power spectrum indicative
of diaphragmatic fatigue have been documented during severe
whole body exercise (61) and during loaded breathing in normal subjects, in female patients during delivery (39), as well as
in ventilator-dependent patients having weaning problems (2).
Methodology. The methodology required for electromyographic assessment is reviewed in detail in EMG EQUIPMENT
AND DATA ANALYSIS in Section 3 of this Statement.
Advantages. Studies of normal subjects have shown a good
correlation between EMG power spectrum shifts and force or
pressure losses at high stimulation frequencies (high-frequency
fatigue) but not with force loss at low stimulation frequencies
(low-frequency fatigue) (62). High-frequency fatigue is typically associated with failure at the neuromuscular junction or at
the sarcolemma. In line with these predictions, a good correspondence has been found for the human diaphragm between
the rate of power spectrum shift, the pressuretime product
(63), and the changes of the shape of the action potential wave
form measured during phrenic nerve stimulation.
Disadvantages. The etiology of power spectral shifts with
fatigue is still controversial. Possible mechanisms include a
slowing of muscle fiber conduction velocity, a widening of the
action potential waveform, a decrease in motor unit discharge
rate, or synchronization of motor units firing (64). None of
these can be directly linked to a fatiguing process at the sarcomere level. Power spectrum shifts are therefore related to central motor control, or reflex pathways, or changes in electrolyte or metabolite concentrations within the muscles.
In addition, power spectrum shifts are rapidly reversed on
rest or with reduced activity even though the muscle may remain in a fatigued state. Power spectrum analysis of an EMG,
Figure 4. Time course of various parameters during an isometric contraction held at 40% of maximal force (task) until failure. From the
top: Maximal force: voluntary maximal electrical supramaximal pulses
of the nerve. The vertical lines are pressure swings obtained by electrical stimulation, showing a progressive loss of maximal force. The time
elapsed from the start until task failure is known as the time limit, or
endurance. The exercise is defined as a fatiguing task, because there
was loss of maximal force. Regaining the ability to develop maximal
force takes a few minutes. Regaining the ability to perform the same
task again, however, takes hours. This panel shows the time course of
the central frequency (CF) of the EMG obtained via surface electrodes
and fast Fourier transforms in the same exercise. Decay of CF is fast,
and is a forewarning of task failure. This parameter is an expression of
membrane potential conduction velocity. Relaxation rate: The time
course of relaxation time (if the contraction is interrupted). Control
values are the same as in a rested muscle. This parameter is linked to
failure at the sarcomere level, mainly related to calcium coupling and
release from troponin. Tetanic force: The decrease in force during an
electrical stimulation at 100 or 20 Hz, and its ensuing rate of recovery.
Recovery from fatigue is faster when the muscle is probed with 100 Hz
than with 20 Hz; the former is proposed to be caused by conduction
mechanisms, whereas the latter is mediated by contraction mechanisms. Spontaneous shortening: Spontaneous shortening of the diaphragm before (100%) and after task failure, and time of recovery.
Figure used by permission from Dr. A. E. Grassino.
therefore, cannot provide an indication as to the state of the

contractile system or the excitationcontraction coupling process, or how these may change with fatigue. As mentioned
previously, because of the close association of these indices
with neural or sarcolemmal events, power spectral shifts recover quickly with rest (within 5 minutes). As a result, these
indices can be markedly affected by the breathing pattern (63)
and the breathing strategy employed (29), which in turn may
cause a high breath-to-breath variability.
Applications. Because of the problems listed above, this
test cannot be taken as a reliable global index of the development of muscle fatigue. The principal utility of this test is that
demonstration of an EMG spectral power shift in a working
muscle may provide some clue to the development of an alteration in neuromuscular transmission.
Muscle Responses to External Stimulation
Pressurefrequency relationships. Rationale. Fatigue, defined

as a decrease in the pressure- or force-generating capacity of a
577
muscle under loaded conditions, can be most specifically detected by recording the pressure or forcefrequency curve of
that muscle in response to artificial motor nerve stimulation.
Of the respiratory muscles, the diaphragm (16, 65) and the
sternomastoid (66) muscles are most amenable to this form of
testing. Low-frequency fatigue has been documented for these
muscles in normal subjects during loaded breathing (16) as
well as during intense exercise (67).
Methodology. The methodology required for phrenic nerve
stimulation and sternomastoid stimulation is reviewed in detail in Section 2 of this Statement.
Advantages. This technique overcomes many of the difficulties associated with volitional or spontaneous breathing efforts. Indeed, the responses are not complicated by possible
variations in the level of effort expended. The response of a
particular muscle can also be studied in isolation, free from
the activity of other muscles. Changes in the shape of the pressure or forcefrequency curves also give indications as to the
underlying mechanism of fatigue. For example, a decreased
pressure or force at high stimulation frequencies may be indicative of impairment at the neuromuscular junction or at the
sarcolemma, whereas a decreased force or pressure at low
stimulation frequencies may suggest a possible impairment of
excitationcontraction coupling.
Disadvantages. This is a difficult test to perform. Tetanic
stimulation can also be painful and it may be necessary to
anesthetize the skin near the electrodes. To overcome this
problem, partial pressurefrequency curves may be constructed by using twin pulses and by varying the intervals between the pulses (68, 69). These are better tolerated than tetanic stimulation and can provide comparable information regarding
the presence of high- and low-frequency fatigue. Because of a
large intersubject variability in the responses to artificial stimulation, fatigue can be reliably detected by these techniques
only when a subject serves as his/her own control. A possible
exception concerns the ratio of the force or pressure developed at a low stimulation frequency (i.e., 20 Hz) over that at a
high stimulation frequency (i.e., 100 Hz), for which some critical value may be recognized below which low-frequency fatigue may be said to be present (16, 65, 66, 68).
Application. Although this test provides a means of directly detecting the development of muscle fatigue, applicability of this approach is limited by (1) patient discomfort associated with high-frequency stimulation, (2) equipment expense
and complexity, and (3) the need to carefully control for variation in body position, lung volume, and the electrodenerve
interface. Advances in magnetic stimulation techniques may
allow a variation of this form of testing to reach more widespread clinical application in the future, but this test is currently limited to research applications.
Single twitch stimulation. Rationale. As an alternative to
tetanic or twin pulse stimulation, recording of muscle twitches
in response to single nerve shocks can be employed to detect
the presence of low-frequency fatigue (68). Twitch responses
are much easier to obtain but are more variable than tetanic
responses and are subject to additional variations caused by
phenomena such as twitch potentiation (70).
Methods and equipment. The methodology required for phrenic
nerve stimulation is reviewed in Section 2 of this Statement.
Advantages. This technique is nonvolitional, eliminating
concerns about patient effort in the interpretation of obtained
results. In addition, because only single twitches are evoked
when employing this technique, much less patient discomfort is
involved when compared with that produced by construction of
the forcefrequency relationship. Because a single shock is, by
definition, a low-frequency stimulus, this approach also pro-
578
vides a means of detecting the development of low-frequency

fatigue, whereas measurement of maximal volitionally produced pressure does not. In addition, the magnitude of the compound action potential evoked during single twitches can be
measured and monitored over time; correlation of this assessment with force generation over time may provide a means of
detecting alterations in neuromuscular transmission.
Disadvantages. The application of these tests has been
largely limited to scientific investigations in normal subjects
and patients with lung (71) or neuromuscular diseases (72).
Their application in clinical settings is more difficult, partly
because of the factors already mentioned but also because of
the many pieces of equipment that are required to perform
these tests. In the case of the diaphragm, these difficulties are
compounded by the necessity of stimulating the phrenic
nerves bilaterally. In addition, it is critical that supramaximality be attained during electrical stimulation for this test to be
useful. Unless electrical current is sufficiently high (i.e., increased to a level 150% of that required to initially attain a
maximal signal), alterations in the compound action potential
over time may simply reflect local, axonal changes. The introduction of magnetic stimulation (73) and of phonomyography
(74) may help overcome some of these difficulties.
Applications. Although this technique holds promise, more
work is needed before this test can be used in the clinical
arena. Because of the technical problems detailed above, it is
difficult to use present stimulation techniques to accurately
evaluate the time course of fatigue, and this approach is currently better suited for evaluation of the effect of an intervention or treatment on muscle function. Once standardization of
this technique is achieved, it probably offers the greatest
promise to provide an objective index of the development of
muscle fatigue.
CONCLUSION
This Section of the Statement has reviewed the complex process of muscle fatigue and has discussed the available direct
and indirect measurements relevant to the assessment of fatigue of the respiratory muscles.
Although a variety of measures and indices have been employed to assess the development of respiratory muscle fatigue in
research, there is no well described technique that has been successfully developed and tested to permit precise identification of
respiratory muscle fatigue in the clinical setting. Of the tests reviewed, breathing pattern analysis and measurement of thoracoabdominal motion are nonspecific indices that do not directly
measure fatigue. Analysis of the pressuretime index provides a
useful conceptual framework, but not a specific test of fatigue.
In the research environment, serial measurement of maximal voluntary respiratory pressures, assessment of maximum
relaxation rates, frequency domain EMG analysis, and measurement of respiratory muscle pressures in response to electrical or magnetic nerve stimulation are all techniques that can be
used to assess the evolution of respiratory muscle fatigue. Of
these techniques, serial measurement of respiratory muscle
pressure generation in response to electrical or magnetic stimulation is arguably the best technique to directly assess the development of respiratory muscle fatigue at the present time, and it
offers the greatest promise for future development into an objective test of respiratory muscle fatigue in the clinical arena.
References
1. Cohen C, Zagelbaum G, Gross D, Roussos C, Macklem PT. Clinical manifestations of inspiratory muscle fatigue. Am J Med 1982;73: 308316.
VOL 166
2002
2. Respiratory Muscle Fatigue Workshop Group. Respiratory muscle fatigue:

NHLBI Workshop Summary. Am Rev Respir Dis 1990;142:474480.
3. Bellemare F, Bigland-Ritchie B. Assessment of human diaphragm
strength and activation using phrenic nerve stimulation. Respir Physiol 1984;58:263277.
4. Scardella A, Santiago T, Edelman N. Naloxone alters the early response to an inspiratory flow-resistive load. J Appl Physiol 1989;67:
17471753.
5. Bellemare F, Bigland-Ritchie B. Central components of fatigue assessed
by phrenic nerve stimulation. J Appl Physiol 1987;62:13071316.
6. Allen GM, Gandevia SC, McKenzie DK. Reliability of measurements of
muscle strength and voluntary activation using twitch interpolation.
Muscle Nerve 1995;18:593600.
7. McKenzie DK, Bigland-Ritchie B, Gorman RB, Gandevia SC. Central
and peripheral fatigue of human diaphragm and limb muscles assessed
by twitch interpolation. J Physiol (Lond) 1992;454:643656.
8. Santiago T, Remolina C, Scoles V, Edelman N. Endorphins and control
of breathing: ability of naloxone to restore the impaired flow-resistive
load compensation in chronic obstructive pulmonary disease. N Engl J
Med 1981;304:11901195.
9. Petrozzino JJ, Scardella A, Santiago T, Edelman N. Dichloracetate
blocks endogenous opioid effects during inspiratory flow-resistive
loading. J Appl Physiol 1992;72:590596.
10. Bellofiore S, DiMaria G, Privitera S, Sapienza S, Milic-Emili J, Misgtretta A. Endogenous opioids modulate the increase in ventilatory
output and dyspnea during severe acute bronchoconstriction. Am Rev
Respir Dis 1990;142:812816.
11. Adams JM, Farkas GA, Rochester DF. Vagal afferents, diaphragm fatigue, and inspiratory resistance in anesthetized dogs. J Appl Physiol
1988;64:22792286.
12. Aldrich TK. Respiratory muscle fatigue. Clin Chest Med 1988;9:225236.
13. Aldrich TK. Transmission failure of the rabbit diaphragm. Respir Physiol 1987;69:307319.
14. Bazzy AR, Donelly DF. Diaphragmatic fatigue during loaded breathing: role of neuromuscular transmission. J Appl Physiol 1993;74:
16791683.
15. Aubier M, Farkas G, De Troyer A, Mozes R, Roussos C. Detection of
diaphragmatic fatigue in man by phrenic stimulation. J Appl Physiol
1981;50:538544.
16. Westerblad H, Allen DG. Changes of myoplasmic calcium concentration during fatigue in single mouse muscle fibers. J Gen Physiol 1991;
98:615635.
17. Westerblad H, Lannergren J, Allen DG. Fatigue of striated muscles:
metabolic aspects. In: Roussos C, editor. The thorax, 2nd ed. New
York: Marcel Dekker; 1995.
18. Nosek TM, Leal-Cardoso JH, McLaughlin M, Godt RE. Inhibitory influence of phosphate and arsenate on contraction of skinned skeletal
and cardiac muscle. Am J Physiol 1990;259:C933C939.
19. Jones DA. Muscle fatigue due to changes beyond the neuromuscular
junction. In: Porter R, Whelan J, editors. Human muscle fatigue: physiological mechanisms. London: Pitman Medical; 1981. p. 178190.
20. Shindoh C, DiMarco A, Thomas A, Manubay P, Supinski G. Effect of
N-acetylcysteine on diaphragm fatigue. J Appl Physiol 1990;68:2107
2113.
21. Anzueto A, Andrade FH, Maxwell LC, Levine SM, Lawrence RA, Gibbons WJ, Jenkinson SG. Resistive breathing activates the glutathione
redox cycle and impairs performance of the rat diaphragm. J Appl
Physiol 1992;72:529534.
22. Moxham J, Wiles CM, Newham DD, Edwards RHT. Contractile function and fatigue. In: Porter R, Whelan J, editors. Human muscle fatigue: physiological mechanisms. London: Pitman Medical; 1981. p.
197205.
23. Wragg S, Aquilina R, Moran J, Hanmegerd C, Green M, Moxham J. Diaphragm fatigue following maximum ventilation in man. Am Rev Respir
Dis 1992;145:A147.
24. Tobin MJ, Perez W, Guenther SM, Semmes BJ, Mador MJ, Allen SJ,
Lodato RF, Dantzker DR. The pattern of breathing during successful
and unsuccessful trials of weaning from mechanical ventilation. Am
Rev Respir Dis 1986;134:11111118.
25. Mador MJ, Tobin MJ. The effect of inspiratory muscle fatigue on
breathing pattern and ventilatory response to CO2. J Physiol (Lond)
1992;455:1732.
26. Roussos C, Fixley M, Gross D, Macklem PT. Fatigue of inspiratory muscles and their synergic behavior. J Appl Physiol 1979;46:897904.
27. Fitting JW, Bradley TD, Easton PA, Lincoln MJ, Goldman MD,
Grassino A. Dissociation between diaphragmatic and rib cage muscle
fatigue. J Appl Physiol 1988;64:959965.

28. Tobin MJ, Perez W, Guenther SM, Lodato RF, Dantzker DR. Does rib
cageabdominal paradox signify respiratory muscle fatigue? J Appl
Physiol 1987;63:851860.
on human diaphragm fatigue. J Appl Physiol 1982;53:11901195.
30. Zocchi L, Fitting JW, Majani U, Fracchia C, Rampulla C, Grassino A.
Effect of pressure and timing of contraction on human rib cage muscle
fatigue. Am Rev Respir Dis 1993;147:857864.
31. Bellemare F, Grassino A. Force reserve of the diaphragm in COPD patients. Appl Physiol 1983;55:815.
32. Vassilakopoulos T, Zakynthinos S, Roussos C. The tensiontime index
and frequencytidal volume ratio are the major pathophysiologic determinants of weaning failure and success. Am J Respir Crit Care Med
1988;158:378385.
33. Bai TR, Rabinovitch J, Pardy RL. Near-maximal voluntary hyperpnea
34. Loke J, Mahler DA, Virgulto JA. Respiratory muscle fatigue after marathon running. J Appl Physiol 1982;52:821824.
35. Chevrolet JC, Tschopp JM, Blanc Y, Rochat T, Junod AF. Alterations
in inspiratory and leg muscle force and recovery pattern after a marathon. Med Sci Sports Exer 1993;25:501507.
36. Nava S, Zanotti E, Ambrosino N, Fracchia C, Scarabelli C, Rampulla C.
Evidence of acute diaphragmatic fatigue in a natural condition: the
diaphragm during labor. Am Rev Respir Dis 1992;146:12261230.
37. Bye PTP, Esau SA, Walley KR, Macklem PT, Pardy RL. Ventilatory
muscles during exercise in air and oxygen in normal men. J Appl
Physiol 1984;56:464471.
38. Miller JM, Moxham J, Green M. The maximal sniff in the assessment of
diaphragm function in man. Clin Sci 1985;69:9196.
39. Laroche CM, Mier AK, Moxham J, Green M. The value of sniff esophageal pressures in the assessment of global inspiratory muscle strength.
40. Hritier F, Rahm F, Pasche P, Fitting JW. Sniff nasal inspiratory pressure: a noninvasive assessment of inspiratory muscle strength. Am J
41. Uldry C, Fitting JW. Maximal values of sniff nasal inspiratory pressure
in healthy subjects. Thorax 1995;50:371375.
42. Esau SA, Bellemare F, Grassino A, Permutt S, Roussos C, Pardy RL.
Appl Physiol 1983;54:13531360.
43. Esau SA, Bye PTP, Pardy RL. Changes in rate of relaxation of sniffs with
44. Koulouris N, Vianna LG, Mulvey DA, Green M, Moxham J. Maximal
relaxation rates of esophageal, nose, and mouth pressures during a
sniff reflect inspiratory muscle fatigue. Am Rev Respir Dis 1989;193:
12131217.
45. Mador MJ, Kufel TJ. Effect of inspiratory muscle fatigue on inspiratory
muscle relaxation rates in healthy subjects. Chest 1992;102:17671773.
46. Kyroussis D, Mills G, Hamnegard CH, Wragg S, Road J, Green M, Moxham J. Inspiratory muscle relaxation rate assessed from sniff nasal
pressure. Thorax 1994;49:11271133.
47. Mulvey DA, Koulouris NG, Elliott MW, Moxham J, Green M. Maximal
relaxation rate of inspiratory muscle can be effort-dependent and reflect
the activation of fast-twitch fibers. Am Rev Respir Dis 1991;144:803806.
48. Kyroussis D, Polkey MI, Keilty SEJ, Mills GH, Hamnegard CH, Moxham J, Green M. Exhaustive exercise slows inspiratory muscle relaxation rate in chronic obstructive pulmonary disease. Am J Respir Crit
Care Med 1996;153:787793.
49. Goldstone JC, Green M, Moxham J. Maximum relaxation rate of the diaphragm during weaning from mechanical ventilation. Thorax 1994;49:5460.
50. Lippold OCJ, Redfearn JWT, Vuco J. The electromyography of fatigue.
Ergonomics 1960;3:121131.
579
51. Stephens JA, Taylor A. Fatigue of maintained voluntary maximal contraction in man. J Physiol (Lond) 1972;220:118.
52. Aubier M, Trippenbach T, Roussos C. Respiratory muscle fatigue during cardiogenic shock. J Appl Physiol 1981;51:499508.
53. Grassino E, Goldman MD, Mead J, Sears TA. Mechanics of the human
diaphragm during voluntary contractions: statics. J Appl Physiol 1978;
44:829839.
54. McKenzie DK, Gandevia SC. Electrical assessment of respiratory muscles.
In: Roussos C, editor. Lung biology in health and disease, 2nd ed., Part
B. Vol. 85: The thorax. New York: Marcel Dekker; 1996. p. 10291048.
55. Gandevia SC, McKenzie DK. Human diaphragmatic EMG: changes
with lung volume and posture during supramaximal phrenic stimulation. J Appl Physiol 1986;60:14201428.
56. Beck J, Sinderby C, Weinberg J, Grassino A. Effects of muscle-to-electrode distance on the human diaphragm electromyogram. J Appl
Physiol 1995;79:975985.
57. Sinderby C, Lindstrom L, Comtois N, Grassino A. Effects of diaphragm
shortening on the mean action potential conduction velocity in canines. J Physiol (Lond) 1996;490:207214.
58. Sinderby CA, Comtois A, Thomson R, Grassino AE. Influence of the bipolar electrode transfer function on the electromyogram power spectrum. Muscle Nerve 1996;19:290301.
59. Sinderby C, Lindstrom L, Grassino AE. Automatic assessment of electromyogram quality. J Appl Physiol 1995;79:18031815.
60. Gross D, Grassino A, Ross WRD, Macklem PT. Electromyogram pattern of diaphragmatic fatigue. J Appl Physiol 1979;46:17.
61. Pardy RL, Bye PTP. Diaphragmatic fatigue in normoxia and hyperoxia.
J Appl Physiol 1985;58:738742.
62. Moxham J, Edwards RHT, Aubier M, De Troyer A, Farkas G, Macklem
PT, Roussos C. Changes in EMG power spectrum (high/low ratio)
with force fatigue in man. J Appl Physiol 1982;53:10941099.
63. Bellemare F, Grassino A. Evaluation of human diaphragm fatigue. J
Appl Physiol 1982;53:11961206.
64. DeLuca CJ. Myoelectric manifestations of localized muscular fatigue in
humans. Crit Rev Biomed Eng 1984;11:251279.
65. Moxham J, Morris AJR, Spiro SG, Edwards RHT, Green M. Contractile
properties and fatigue of the diaphragm in man. Thorax 1981;36:164168.
66. Moxham J, Wiles CM, Newham D, Edwards RHT. Sternomastoid muscle function and fatigue in man. Clin Sci Mol Med 1980;59:463468.
67. Johnson BD, Babcock MA, Sumanand OE, Dempsey JA. Exercise induced
diaphragmatic fatigue in healthy humans. J Appl Physiol 1993;460:385405.
68. Yan S, Gauthier AP, Similowski T, Faltus R, Macklem PT, Bellemare F.
Forcefrequency relationships of in vivo human and in vitro diaphragm using paired stimuli. Eur Respir J 1993;6:211218.
69. Polkey MI, Kyroussis D, Hamnegard CH, Hughes PD, Rafferty GF,
Moxham J, Green M. Paired phrenic nerve stimuli for the detection of
diaphragm faigue in humans, Eur Respir J 1997;10:18591864.
70. Wragg S, Hamnegard C, Road J, Kyroussis D, Moran J, Green M. Potentiation of diaphragmatic twitch after voluntary contraction in normal subjects. Thorax 1994;49:12341237.
71. Similowski T, Yan S, Gauthier AP, Macklem PT, Bellemare F. Contractile properties of the human diaphragm during chronic hyperinflation.
N Engl J Med 1991;325:917923.
72. Mier-Jedrzejowicz A, Brophy C, Moxham J, Green M. Assessment of diaphragm weakness. Am Rev Respir Dis 1988;137:877883.
67:13111318.
74. Petitjean M, Bellemare F. Phonomyogram of the diaphragm during unilateral and bilateral phrenic nerve stimulation and changes with fatigue. Muscle Nerve 1994;17:12011209.
6. Assessment of Chest Wall Function

Mechanical properties and functions of the chest wall can be
assessed by measurements of lung volume displacement, chest
wall motion, and pressures measured at various locations in
the chest wall. Respiratory muscle activation can be further
characterized by electromyography (EMG). Techniques of
pressure measurement are presented in Section 2 of this Statement. The framework for interpreting pressures in the chest
wall is presented in this article.
A pressure difference between two points may characterize

two or more different structures or groups of structures. For
example, the pressure difference between the pleural space
and the body surface in a spontaneously breathing person is
both the transthoracic (transchest wall pressure, Pcw) and the
negative of transpulmonary pressure (PL).
PRESSURES IN THE CHEST WALL
Scientific Basis
In respiratory mechanics, it is important to distinguish between the two uses of the word pressure. In one case it denotes a pressure measured at a given location, as in pleural
pressure. In the other case it denotes a difference in pressure
between two points, usually on opposite sides of a structure,
such as transpulmonary pressure, defined as the difference
between pressure at the airway opening (Pao) and pressure in
the pleural space (Ppl). Pressures are usually measured relative to barometric pressure (i.e., they are taken to be zero
when they are equal to barometric pressure).
Pressures at a point are usually assumed to be representative of the pressure in that space (see Figure 1 in Section 2 of
this Statement). This simplification must be qualified when
variations of pressure within a space are to be expected (1). In
particular, gravity causes vertical gradients in pressure related
to the density of the semisolid or liquid contents of a space: in
the thorax, this gradient is approximately 0.2 cm H2O/cm height
and is affected by lung density; in the abdomen, the gradient is
nearly 1 cm H2O/cm height. Temporal fluctuations in pressure, as in tidal breathing, are little affected by gravitational
gradients. Shear stress resulting from the deformation of elastic, shape-stable organs can cause local variations in pressure,
such as those that occur just below the diaphragm when it displaces the liver during a large forceful contraction (2).
Pressure differences across structures, as opposed to pressures measured at a point, are relevant for characterizing those
structures. The schematic drawing in Figure 1 of Section 1 of
this Statement shows relationships among locations where
pressures can be measured (within circles) and intervening respiratory structures and equipment (within rectangles). Pleural
and abdominal pressures are usually estimated by measuring
esophageal and gastric pressures (Pes and Pga), respectively.
Table 1 and Figure 1 in Section 2 of this Statement list pressures measured at a point and pressure differences across
structures. These differences are usually taken in a direction
such that positive pressure differences expand the structure
(e.g., lung). An exception to this rule is transdiaphragmatic
pressure (Pdi), which has been defined both as pleural pressure minus abdominal pressure, Pdi Ppl Pab, and as its reverse, Pdi Pab Ppl. The complicating effects of gravity
must be considered when pleural pressure is estimated from
esophageal pressure (Pes) and abdominal pressure is estimated from gastric pressure (Pga). When the diaphragm itself
is completely relaxed and the actual pressure difference across
the diaphragm is nil, the measured transdiaphragmatic pressure has a minimum value, usually approximately 10 cm H2O,
which is attributed mostly to the gravitational hydrostatic difference between esophageal and gastric pressures. This hydrostatic
transdiaphragmatic pressure, which changes only
slightly
with breathing (3), is usually subtracted from reported
DOI: 10.1164/rccm.AT0602
measurements
of Pdi.
Pressure differences across viscoelastic, plastoelastic structures such as the lungs and chest wall depend on the structures volume, volume history, and rate of change of volume.
Accordingly, pressure differences across respiratory structures
are often represented as characteristic pressurevolume (PV)
curves (4). For the relaxed respiratory system, transpulmonary, transthoracic, and transrespiratory pressures are usually
plotted against lung volume in a Rahn diagram (Figure 1). The
PV characteristics shown in Figure 1 are of a relaxed subject
slowly inflated or deflated by a pressure source at the airway
opening. All the passive structures show an increase in volume
with an increase in the pressure difference across them. When
two pneumatic structures are in series, for example, the lung
and the chest wall, the pressure difference across both structures (the transrespiratory pressure) is the sum of the pressure
differences across each, and the volume displacements of the
whole are equal to the volume displacements of each part. The
PV curve can be locally described by the volume at a given
pressure and the slope (compliance) at that point. The compliance of a passive structure at a given volume is the ratio of volume change to pressure change (i.e., the slope of the characteristic PV curve at that volume).
ASSESSMENT OF THE PROPERTIES OF THE RELAXED

HUMAN CHEST WALL: RAHN DIAGRAM
Methodology
Lung volume displacements and pressures are measured as

described in Sections 1 and 2 of this Statement. The following
properties of the chest wall and lung are found by analysis of
Rahn diagrams:
1. In trained subjects, the static PV characteristic of lung compliance is obtained from transpulmonary pressure (PL
Pao Pes) during an interrupted exhalation from total
lung capacity (TLC) to residual volume (RV) with the glottis held open. The quasistatic deflation curve is similar, and
is measured during a slow exhalation (expiring at flows less
than 0.3 L/second). For subjects who cannot satisfactorily
perform these maneuvers, PL can be measured during intermittent airway occlusions, 24 seconds long, in an expiration from TLC to RV, or during interrupted deflation of
the respiratory system with a supersyringe, valve, or other
device.
2. The PV characteristic of the relaxed chest wall (Pcw) is obtained from esophageal pressure (Pes) during a slow relaxed exhalation through pursed lips or other high resistance from TLC to functional residual capacity (FRC) and
during passive inhalation with relaxation against an intermittently occluded airway between RV and FRC. Alternatively, it can be measured during exhalation from TLC to
RV with periodic airway occlusions with relaxation. Relaxation above FRC, however, may be difficult for untrained
subjects, and relaxation below FRC can usually be achieved
581
only by highly trained subjects. Normal compliance at FRC

is approximately 0.2 L/cm H2O.
3. The PV characteristic of the relaxed respiratory system is
obtained by plotting Pao versus VL during the maneuvers
described above (Prs Pao Pbs [body surface pressure]). Normal values are approximately 0.1 L/cm H2O at
volumes of 4060% of the vital capacity (VC).
Advantages
The Rahn diagram is useful for describing the elastic properties of passive systems. Each curve reflects the pressure difference developed by this structure for a range of volumes. Static
compliance determined from these curves can be used for diagnosis. For example, the compliance of the lung is decreased
in interstitial lung disease and increased in emphysema. Chest
wall compliance is decreased in ankylosing spondylitis and
obesity.
Disadvantages
Whereas the lungs PV characteristic (i.e., the plot of elastic

recoil pressure of the lung, PL,el, versus lung volume) is relatively easy to obtain in untrained subjects, the elastic recoil
pressure of the chest wall (Pcw,el versus VL) is difficult to
measure because it requires complete relaxation of the respiratory muscles at various lung volumes. Relaxation can be
monitored via surface EMG of the chest wall. Failure to relax
the respiratory muscles is also revealed when repeated measurements of passive PV curves of the chest wall are not identical. In subjects who can achieve relaxation at FRC but not at
volumes above FRC, a relaxation curve can be approximated
by extending a line from the relaxation point, assuming a normal chest wall compliance of 0.1 L/cm H2O. The chest wall
compliance can also be estimated in untrained subjects by the
weighted spirometer technique (5). These estimations may, however, be unreliable in hyperinflated patients (e.g., in chronic
obstructive pulmonary disease) who never reach a true (static)
relaxation volume even when they are relaxed during exhalation. This is a well-established test, although it is seldom used
for clinical diagnosis.
Figure 1. The Rahn diagram shows the pressurevolume relationships

of the relaxed chest wall (Pcw), lungs (PL), and the total respiratory
system (Prs). Ordinate: Lung volume expressed as a percentage of vital
capacity (VC). Abscissa: Pressures, each defined by the respective
equations. Horizontal dashed line: Relaxation volume of the respiratory
system. Pes esophageal pressure; Pbs body surface pressure; Pao
airway opening pressure. Modified from Campbell EJM, Agostoni E,
Davis JN, editors. The respiratory muscles: mechanics and neural control, 2nd ed. Philadelphia, PA: W.B. Saunders; 1970. p. 55.
ASSESSMENT OF THE FUNCTION OF THE ACTIVE CHEST

WALL: CAMPBELL DIAGRAM
Scientific Basis
To evaluate respiratory muscle action, a Campbell diagram, in

which lung volume on the ordinate is plotted against pleural
pressure on the abscissa, can be constructed. In this diagram,
pleural pressure has differing significance depending on the
maneuver. Consider a subject who is slowly inflated and deflated passively by a syringe connected to the airway while the
respiratory muscles are relaxed (passive inflation; Figure 2).
The pleural pressure, which is equal to transthoracic pressure,
rises and falls, describing the characteristic PV curve for the
relaxed chest wall, which is the same as that in the Rahn diagram (Figure 1). Alternatively, during active slow inhalation
and exhalation with an open glottis, the pleural pressure (in
this case equal to transpulmonary pressure with a negative
sign) becomes more subatmospheric as the lungs inflate (active inflation; Figure 2), describing the lungs characteristic PV
curve, which appears as a mirror image of the lungs curve in
Figure 1. In the Campbell diagram the two curves intersect at
relaxation volume at a pleural pressure of about 5 cm H2O.
The intersection represents the equal and opposite elastic recoils of the lung and chest wall.
During inhalation, the pleural pressure is the pressure
across the active chest wall. The pressure generated by the inspiratory muscles is simply the pressure difference between
the active chest wall characteristic and the relaxed chest wall
characteristic. Work done by the inspiratory muscles,
Pmus dv,
is represented by the hatched area in Figure 2. The horizontally hatched area represents the work done to overcome elastic recoil of the lung and chest wall. Additional pressure is necessary to overcome airway resistance and lung tissue resistance;
this work is shown with vertical hatching. Total work is therefore the sum of elastic and resistive work per inhalation, and is
usually multiplied by breathing frequency and expressed as g
cm/ml. Work of breathing was found to average 2.2 0.92 g
cm/ml at a respiration frequency of 15 breaths/minute (6), and
was independent of age or sex. This is a test of great physiological interest and is widely used in research. It is seldom
used for clinical evaluations. Reference 6 gives a full account
of the complexities of how the work is done by the coordination of the various respiratory muscles.
Figure 3 shows a Campbell diagram with the addition of
maximal static inspiratory pressure (MIP) and maximal expiratory pressure (MEP) during efforts against an occlusion
(outer dashed lines). The MIP is greatest (most subatmospheric) at low lung volumes, whereas the MEP is greatest
(most positive) at high lung volumes, largely because of the
lengthtension characteristics of the inspiratory and expiratory muscles. At high lung volume, the diaphragm and other
inspiratory muscles are shorter, whereas expiratory muscles
are longer (4). The inner dashed and dotted line in Figure 3 indicates the pleural pressure required to balance the elastic recoil of the lungs. TLC is at the intersection of these lines,
where maximal inspiratory pleural pressure is balanced by the
lungs elastic recoil. The innermost loop represents resting
breathing at FRC, as is shown for inspiration only in Figure 3.
The pressurevolume relationship during a maximal forced
inspiration and expiration are shown as the inner solid line
loop in Figure 3. Pressures at every volume are reduced from
maximal static pressures because the muscle is shortening. The
loss of maximal inspiratory pressure (the difference in pressure
between the dashed and solid lines at a given lung volume in
582
Figure 2. Campbell diagram. Graphical analysis of the work done during a breathing cycle by the inspiratory muscles. Vertical hatching:
Work done to overcome flow resistance of the lungs. Horizontal hatching: Work done to overcome elastance of the lungs and chest wall.
Modified by permission from Macklem PT, Mead J, editors. Handbook
of physiology. Vol. 3: The respiratory system, Part 3. Bethesda, MD:
American Physiological Society; 1986. p. 495.
Figure 3) was estimated to be approximately 7%/L/second of

flow at a volume of FRC 1 L (7). The muscle pressure at this
volume represents the maximal capacity (Pcap) of the inspiratory muscle to generate pressure while shortening maximally.
During submaximal exercise, for example, healthy subjects
may require esophageal pressures in the range of 30 cm H2O
to ventilate the lungs, or approximately 40% of Pcap. When
pressures in that range are achieved while breathing for several
minutes with a duty cycle of 0.5, muscle fatigue may result (see
Section 5 of this Statement). At higher levels of ventilation
(maximal voluntary ventilation), at which peak flow can reach
up to 10 L/second, there is a decrease in maximal inspiratory
muscle pressure within 1520 seconds, attributed to fatigue.
VOL 166
2002
Figure 3. Campbell diagram with reduced scale to show maximal

static inspiratory pressures (MIP) and maximal expiratory pressures
(MEP) at various lung volumes (dashed lines), pressures during maximal dynamic inspiration and expiration in a forced vital capacity (FVC)
maneuver (outermost loop), and spontaneous breathing at rest (innermost loop).
cles. The relationship between work and oxygen consumption

makes it possible to calculate the efficiency of respiratory
muscles (see Figure 3 in Section 4 of this Statement). The difference between maximal static inspiratory pressures and
peak actual pleural pressure measured during breathing indicates the muscle force reserve, an index that helps assess the
likelihood of fatigue. This is a well-established method, often
used in clinical research.
Disadvantages
To infer respiratory muscle action, the pleural pressure must

be referred to the PV curve of the relaxed chest wall as in the
example above. For example, a given positive pleural pressure
may reflect either expiratory muscle activity at FRC or slight
inspiratory muscle activity at a high lung volume. Furthermore, difficulties in measuring the relaxation characteristic in
untrained subjects can make estimates of Pmus uncertain. Although the Campbell diagram is easy to plot, its interpretation
requires some practice.
Equipment
Measurements of pressure and volume are described in Sections 1 and 2 of this Statement.
ESTIMATION OF VENTILATION BASED ON CHEST WALL

MOTION: KONNOMEAD DIAGRAM
Applications
Scientific Basis
The following properties of the chest wall and lung are found
by analysis of Campbell diagrams:
Quantitative measurements of respiratory system motion are

usually based on measurements of lung volume and displacements of chest wall structures, including the abdominal wall.
Because the tissues of the chest wall are essentially incompressible, volume changes (displacements) of the chest wall surfaces are nearly equal to volume changes of the lungs and can
be used to noninvasively estimate lung volume, without the
need for a mouthpiece, mask, or other connection to the airway. (Changes in intrathoracic blood volume cause differences
between chest wall and lung volume displacements, but these
are usually negligible.) The relevance of noninvasive measurements of ventilation is shown in BREATHING PATTERNS in Section 10 of this Statement.
During quiet breathing in a subject at rest, the surface of
the chest wall moves in a predictable way as lung volume increases and decreases. Various dimensions of the chest wall
have been measured to estimate changes in chest wall (or lung)
volume. The chest pneumogram is a record of changes in
thoracic circumference that provides a qualitative measure of
ventilation. However, because the major compartments of the
chest wall, the rib cage and abdomen, move independently in
most conscious subjects, measuring overall chest wall displacement accurately usually requires two or more simultaneous
measurements of displacement.
1. The quasistatic (or static) PV characteristic of the lung is

obtained from esophageal pressure (Pes PL,el) during
a slow (or halting) inhalation and exhalation with the glottis held open.
2. The PV characteristic of the relaxed chest wall is described
above as in the Rahn diagram.
3. Pressure generated by respiratory muscles (Pmus) is the
horizontal distance (i.e., change in pressure) between the
relaxation characteristic of the passive chest wall and the
active one.
4. The Campbell diagram is also used to depict the maximal
static inspiratory pressure (MIP) and maximal static expiratory pressure (MEP) measured with the airway occluded.
The values of MIP in healthy young males are shown as the
dashed outer loop in Figure 3. Section 2 of this Statement
discusses the technique used to perform the MIP test and
gives normal values in health and disease.
Advantages
The Campbell diagram is a convenient tool for calculating the

elastic and resistive work of inspiratory and expiratory mus-
583
Konno and Mead (8) established that chest wall volume

change could be assessed by measuring displacements of rib
cage and abdominal surfaces. In their subjects, who were standing still, a single anteroposterior diameter sufficed to indicate
motion of the rib cage; this was also true of the abdomen.
Thus, the chest wall can be described as having two principal
degrees of freedom of motion: Overall chest wall displacements
can be specified by knowing the rib cage displacement and the
abdominal displacement. The two-compartment model of the
chest wall introduced by Konno and Mead and the plot of rib
cage displacement versus abdominal displacement (the Konno
Mead diagram) have been used in numerous studies.
Figure 4 shows a KonnoMead diagram of a subject breathing
quietly and performing an isovolume maneuver. During quiet
inspiration, the rib cage and abdomen move out synchronously, following the rib cageabdomen relaxation characteristic. In the isovolume maneuver, the subject voluntarily shifts
volume between rib cage and abdominal compartments by
contracting and relaxing abdominal muscles with the glottis
closed. Because lung volume is constant, the decrease in abdominal volume (i.e., the volume displaced by inward movement of the abdominal wall) must be equal to the increase in
rib cage volume; two isovolume maneuvers performed at
known lung volumes allow calibration of rib cage and abdominal displacements in terms of lung volume change.
Most methods of measurement of thoracoabdominal displacement require that the subject maintain a constant posture, as spinal flexion affects the relationships among lung volume, rib cage, and abdominal displacements (9). Therefore,
postural change should, as a rule, be minimized. However, it is
possible to measure spinal flexion in addition to rib cage and
abdominal displacements to estimate lung volume in subjects
whose posture changes (10). This method is widely used in
clinical research to assess tidal volume and the relative displacement of rib cagediaphragm.
Disadvantages. Conversely, when thoracic shape distortions

occur, the respiratory magnetometer becomes less accurate in
measuring overall chest wall volume. Whereas during quiet
breathing, one rib cage and one abdominal signal are adequate to describe chest wall displacements and lung volume
changes, during forceful or unusual respiratory efforts, changes
occur in the ellipticity of the rib cage cross-section and in the
relative displacements of cephalic and caudal regions of the
rib cage, making a single diameter inadequate for assessment
of volume displacement. In such circumstances, the addition
of a transverse diameter measurement can improve volumetric accuracy (13).
In all applications, the orientation of the electromagnetic
coils should be controlled to keep the axes of paired coils parallel to each other to avoid errors caused by pitch and yaw. For
transverse diameter measurements, in which the sides of the rib
cage are not parallel, the coils may be fixed to lightweight calipers and held against the body wall by adhesive tape to maintain coil orientation (13). Displacements of the integument and
soft tissues can introduce artifactual signals in obese subjects at
rest and in all subjects during running, moving in bed, and so
on, and thus the devices are less useful in these cases.
Methodology: Respiratory Inductive Plethysmograph
Another device that has gained acceptance since its introduction more than 15 years ago is the respiratory inductive
plethysmograph (RIP), which uses two elastic bands that surround the rib cage and abdomen to provide a signal that varies
with cross-sectional area. The RIP can measure ventilation
within about 5% compared with the spirometer, and can reveal the relative contributions of rib cage and abdomen to
breathing. This device is widely used to monitor ventilation.
DEVICES USED TO MONITOR BREATHING:

PNEUMOGRAPH, MAGNETOMETER, AND RESPIRATORY
INDUCTIVE PLETHYSMOGRAPH
Methodology: Pneumograph
Chest pneumographs are strain gauges that measure thoracic

circumference. They can be constructed simply from a bellows
that generates pressure when stretched, or an elastic tube containing mercury that changes electrical resistance when stretched.
They are used principally for qualitative measurements, for
example, breathing frequency.
Advantages. They are inexpensive, and useful as qualitative measures of ventilation.
Disadvantages. These devices are not accurate in subjects
who are moving or whose breathing movements are not stereotypical.
Methodology: Respiratory Magnetometer
The first device to be used routinely for quantitative measurements of chest wall displacements was the respiratory magnetometer (11), which uses pairs of small electromagnetic coils
fixed to the skin to measure anteroposterior or other diameters of rib cage or abdomen.
Advantages. These electronic calipers are precise, accurate, and consistent, allowing repeated measurements in an individual over many days (12). Also, because they measure diameters (intercoil distance), they are useful for documenting
distortions of the chest wall shape as during asthma attacks
(13) or forceful respiratory efforts (14). They are also useful
for measuring ventilation at rest.
Figure 4. KonnoMead diagram. Ordinate: Displacement of the rib

cage as measured by magnetometers or RIP. Abscissa: Displacement of
the abdominal wall. FRC functional residual capacity, relaxed; TLC
total lung capacity, relaxed against closed airway. TLCFRC line indicates slow expiration against resistance, with muscles relaxed. Isovolume indicates the changes in configuration by voluntarily contracting
the abdominal muscles while relaxing the rib cage with the glottis
closed. 1L Volume separating two isovolume lines. The volume/distance ratio allows calibration. Each point within the diagram represents a unique configuration (muscle length, diaphragm curvature) of
the inspiratory muscles. Loops indicate a tidal breath taken from FRC,
mainly by using rib cage muscles, mainly diaphragm, or both. Filled circles: Three configurations at which static inspiratory efforts were made
as shown in Figure 7 (see INFERRING DIAPHRAGM ACTIVATION AND ELECTROMECHANICAL EFFECTIVENESS FROM EMG).
584
Advantages. The RIP is easy to use. Because the belts encircle a large part of the rib cage and abdomen, integumental
mobility and cross-sectional shape distortions are less problematic than with magnetometers, and therefore the RIP is often
used during sleep or exercise. The RIP is more accurate than
magnetometers in estimating lung volume change (15), perhaps
because its signal varies with cross-sectional area and samples a
larger part of the moving chest wall than do magnetometers.
Disadvantages. The RIP measures changes relative to an unknown baseline, and its accuracy depends on lung volume calibration at the time of study (see below). In addition, movement
of the belts or changing body position can affect calibration.
Calibration of RIP and magnetometers. Numerous methods
for calibrating the magnetometer and RIP signals for estimating lung volume displacement have been described. All methods determine coefficients for rib cage and abdominal signals,
and most require simultaneous spirometric measurements of
lung volume change during periods of breathing in which the
contributions of rib cage and abdomen to tidal volume vary. In
the method described by Konno and Mead (Figure 4), the volumemotion coefficients of rib cage and abdomen are established by having subjects perform an isovolume maneuver at
two volumes that differ by a known amount. Volume calibration is accomplished by extending the isovolume lines to the
axes, where they indicate the known volume difference. Calibration can also be accomplished by using statistical techniques based on tidal breathing without isovolume maneuvers,
and semiautomated procedures and computer programs incorporated into some commercial devices can simplify calibration in practice. Alternatively, one can assume a standard ratio of the volumemotion coefficients of rib cage and abdomen
instead of finding the ratio experimentally, and calibrating the
combined signal with a spirometer (16). The use of a standard
gain ratio when measuring tidal volume is probably as accurate as more elaborate methods of calibration and, importantly, avoids extreme errors in calibration that can occur with
untrained subjects. We recommend use of standard ratios for
measuring tidal volume, especially in untrained subjects.
Methodology: Respiratory Area Flux Meter
Another device, the respiratory area flux meter, also uses expandable coils that surround the rib cage and abdomen, but in
this case the subject must also lie within large fixed-field coils
mounted in a frame (17).
Advantages. The principal advantage of the area flux meter
is that it measures the cross-sectional areas of the rib cage and
abdomen accurately.
Disadvantages. Although promising, this technique has not
yet been widely applied. The necessity for large field coils may
relegate this device to research applications.
OPTICAL DEVICES USED TO MEASURE

CHEST WALL MOTION
A variety of optical techniques using lasers or small lamps and
multiple video cameras can track the three-dimensional position of numerous points on the torso with high precision and
temporal resolution. Optical reflectance motion analysis is
one of these methods (18).
Advantages
These techniques can measure displacements accurately for

estimation of lung volume change and can provide detailed kinematic data for chest wall mechanics (see below). They are
excellent for studying distortion of the chest wall.
VOL 166
2002
Disadvantages
Like other techniques described above, optical techniques are

subject to artifacts from soft tissue movement. Most optical
techniques are expensive and time-consuming, so they have
been limited to research.
INFERRING RESPIRATORY MUSCLE CONTRIBUTION TO

BREATHING FROM CHEST WALL MOTION
Scientific Basis
Thoracoabdominal displacements such as those displayed on a

KonnoMead diagram demonstrate the action, dysfunction,
or paralysis of specific respiratory muscles. For example, with
diaphragmatic paralysis there may be paradoxical inward
motion of the abdominal wall during early inhalation (Figure
4, mainly rib cage). As the active rib cage expands, pleural
pressure decreases, and without diaphragmatic contraction pleural pressure is transmitted to the abdominal compartment,
where it causes a passive inward movement of the abdominal
wall. Likewise, rib cage muscle paralysis causes characteristic
paradoxical movement of the rib cage during inhalation.
Patients with lung disease often have abnormal chest wall
motions, as well. In patients with chronic obstructive pulmonary disease, alternation between rib cage breathing and abdominal breathing (respiratory alternans) and paradoxical
motion of the abdomen during inspiration can herald inspiratory muscle fatigue and impending respiratory failure (19).
This is not, however, pathognomonic of fatigue, because it can
be elicited voluntarily. Furthermore, abdominal pain can trigger such patterns after upper abdominal surgery.
The KonnoMead diagram is only one possible motion
motion diagram of the chest wall, and other plots such as anteroposteriortransverse diameter plots can give additional information about specific respiratory muscle use and describe
distortions of rib cage shape that occur with contraction of
specific respiratory muscles. For example, inspiratory efforts
made principally with the cervical accessory muscles of inspiration can cause the lower rib cage cross-section to become
more elliptical (19). Compared with normal breathing, contraction of the diaphragm by itself causes predominance of abdominal wall over rib cage displacement, increased ellipticity
of the lower rib cage, and predominance of lower over upper
rib cage anteroposterior displacement (16).
Advantages
Chest wall motion is easy to measure and provides information about respiratory muscle activity.
Disadvantages
Motionmotion plots such as KonnoMead diagrams, when

used by themselves to infer specific muscle action, can be ambiguous, because a given motion may be produced by several
different muscular actions. For example, paradoxical motion
of the abdominal wall in inspiration is not necessarily an indication of diaphragm paralysis because of the influence of the
zone of apposition (20) on abdominal displacements (see subsequent passages). When the rib cage expands in a normal inhalation, the outward displacement of the lower rib cage, which
is part of the abdominal container, tends to lower abdominal
pressure and draw the abdominal wall inward even when the
diaphragm is shortening normally (21) Therefore, small paradoxical inward motions of the abdominal wall during inhalation do not necessarily indicate diaphragmatic paralysis (Figure 4). Conversely, diaphragm paralysis may be missed if one
uses only motions to infer muscle action. The paradoxical in-
585
ward motion of the abdominal wall expected during inhalation

with a paralyzed diaphragm may be abolished if inspiration
occurs with simultaneous relaxation of abdominal muscles
that were activated during expiration (22). Unambiguous evidence of muscle action is best achieved by combining displacement and pressure measurements.
Actions of specific rib cage muscles are difficult to infer
from rib cage motions, because numerous muscles have specific actions at numerous sites on the rib cage. Furthermore,
the simultaneous activity of inspiratory and expiratory muscles can obscure effects of individual muscles. Therefore, assessing the actions of specific rib cage muscles usually involves
electromyography and measurements of several rib cage dimensions and pressures (see INFERRING RESPIRATORY MUSCLE
CONTRIBUTION TO BREATHING FROM PRESSUREVOLUME RELATIONSHIPS and INFERRING DIAPHRAGM ACTIVATION AND ELECTROMECHANICAL EFFECTIVENESS FROM EMG).
INFERRING RESPIRATORY MUSCLE CONTRIBUTION TO

BREATHING FROM THE ESOPHAGEALGASTRIC
PRESSURE RELATIONSHIP: MACKLEM DIAGRAM
Scientific Basis
Decramer and Macklem introduced a method for inferring respiratory muscle action by measuring esophageal and gastric
pressures (23) For example, inhalations made with rib cage
muscles alone (as in diaphragm paralysis) result in decreases
in both esophageal pressure, which is normal, and in gastric
pressure, which is not, whereas diaphragmatic inhalations result in increased gastric pressure and a negative swing in Pes
(lung inflation) (Figure 5).
Advantages
This technique is useful and informative when esophageal and

abdominal pressures are recorded.
Disadvantages
Using chest wall pressures by themselves to infer muscle actions leaves some uncertainty, as does the use of displacements
already described here. Macklems method of analysis assumes that abdominal muscles are relaxed. When they are not,
pressure changes are ambiguous and difficult to interpret.
Nevertheless, this remains an interesting research tool and has
potential for development.
tends to displace the abdominal wall inward and increase abdominal pressure. Therefore, a movement down and to the
right of the relaxation line in Figure 6 reliably indicates abdominal muscle contraction. Similarly, diaphragm contraction
(or paralysis) can be inferred from deviations of the diaphragms PV data in relation to its relaxation characteristic. If
the diaphragm shortens or remains relatively isometric and
transdiaphragmatic pressure increases, the diaphragm must be
actively contracting. For examples of the use of such plots, see
Goldman and coworkers (25) and Mead and coworkers (26).
The rib cage presents a challenge in pressurevolume analysis because no single pressure difference characterizes the
pressures acting on the whole structure. The internal pressure
applied to the lung-apposed surface of the rib cage can be
characterized by pleural (esophageal) pressure, but pressure
in the caudal rib cage apposed to abdominal viscera and diaphragm (the zone of apposition) is closer to abdominal (gastric) pressure than pleural pressure. In addition, the diaphragm itself can have an inspiratory effect on the lower rib
cage when it contracts. Therefore, the pressure acting on or
contributed by the rib cage is generally neither pleural nor abdominal pressure but some combination of the two (27). This
complication can be handled by more elaborate analysis of
chest wall motion and pressures (28).
Advantages
PV data can provide unambiguous evidence of diaphragm and

abdominal muscle activity and, with careful analysis, evidence
of rib cage muscle inspiratory and expiratory action.
Disadvantages
Inferring contributions of specific rib cage muscles generally

requires the addition of electromyographic data and a more
complex analysis. The method remains, however, a powerful
research tool, which can be perfected.
INFERRING DIAPHRAGM ACTIVATION AND

ELECTROMECHANICAL EFFECTIVENESS FROM EMG
The Rahn, Campbell, KonnoMead, and Macklem diagrams
allow the inference of respiratory muscle activity via departures from the relaxed pressurevolume, motionrelaxation,
or pressurepressure relationships. The major limitation of
INFERRING RESPIRATORY MUSCLE

CONTRIBUTION TO BREATHING FROM
PRESSUREVOLUME RELATIONSHIPS
Scientific Basis
Pressurevolume measurements of rib cage and abdominal

compartments can provide the basis for mechanical analysis of
the diaphragm, rib cage, and abdominal muscles, allowing inferences about which muscles contribute to a particular breath
or respiratory maneuver. Whereas the Campbell diagram can
be used to infer inspiratory and expiratory activity of all the
respiratory muscles, other PV diagrams can be used to infer
action of specific respiratory muscles. Pressurevolume diagrams of rib cage and abdomen were introduced by Konno
and Mead (24), who showed their respective relaxation characteristics. By comparing pressurevolume data during breathing
with those obtained during relaxation, these diagrams can be
used to infer the action of specific respiratory muscles when
pressure data alone or volume data alone could be misleading.
An example of the pressurevolume diagram for the abdomen
is shown in Figure 6. Contraction of the abdominal muscles
Figure 5. Macklem diagram. The pleural pressureabdominal pressure

relationship for a variety of breaths is shown, emphasizing the use of certain chest wall muscles: mainly rib cage (RC) muscles (upper girdle, intercostals), normal breath, or mainly diaphragm contraction. Dotted
lines: Iso Pdi isopleths show loci of constant transdiaphragmatic pressure.
586
these methods is that both esophageal and gastric pressures

are affected by the action of many muscles, and the analysis of
their individual actions is indirect and qualitative. Relaxation is often difficult to obtain.
The use of the electromyogram (EMG) displayed in the time
domain allows the evaluation of the level and timing of activation of an individual muscle (2831) The frequency domain
analysis of the EMG allows the evaluation of center frequency
or centroid parameter sensitive to changes in the velocity of
conduction of membrane potentials. Center frequency decreases when the muscle fatigues (32, 33) (and see FREQUENCY
DOMAIN ANALYSIS in Section 3 of this Statement).
Simultaneous measurement of integrated EMG of the diaphragm (Edi), the Pdi, and muscle configuration (length) allows
the estimation of activation pressure ratio, as shown in Figure 7
(28, 31, 34). The strength of Edi can be measured as the root
mean square (RMS) (34). Because absolute RMS values are subjected to many variables, normalization is done to the maximal
voluntary RMS obtained by holding an inspiration at TLC. This
test could be used to evaluate the relative activation during resting breathing. It was found that the percentage of RMS swings
measured during resting breathing was 8% of maximal in healthy
subjects and 43% in chronic obstructive pulmonary disease,
whereas there were nonsignificant differences in Pdi swings (34).
From studies of normal subjects, it was established that the
RMS, expressed as a percentage of its maximal activation at
TLC, varies as a function of chest wall configuration as measured by a KonnoMead diagram (28, 31). The RMS/Pdi ratio
has the highest slope at lung volumes near TLC (short muscle)
and smaller slopes at low lung volumes, at which the diaphragm
is contracting against active abdominal muscles. The force of
the diaphragm (for a given percentage of maximum RMS) is
strongly dependent on its length and configuration (Figure 7).
VOL 166
2002
displacements, chest wall motion, and respiratory pressures.

Important techniques include the following:
Assessment of the passive properties of the chest wall (Rahn
diagram): the chest wall is a contractile musculoskeletal
structure that changes volume to accommodate the ventilatory function of the lung. Its passive characteristics are those
of an elastic body, and have been defined by Rahn diagram
(the relationship between the passive relaxation pressure at
the closed mouth and lung volume). The relaxation characteristics of the lung are measured by the relationship between
volume and esophageal pressure. By subtraction, the passive relaxation properties of the chest wall are calculated.
The Rahn diagram is of considerable conceptual importance,
although seldom constructed outside a research context.
Assessment of the active chest wall (Campbell diagram): lung
volume is plotted against pleural pressure and the diagram
permits calculation of respiratory muscle work, both elastic
and resistive. The measurement of work and oxygen consumption helps calculate muscle efficiency. If pressure data
from maximum efforts against a closed airway (PI,max,
PE,max), across the range of lung volume, are plotted on the
Campbell diagram a useful comparison can be made with
the pressures achieved during maximum dynamic inspiration and expiration or resting ventilation.
Estimation of ventilation by chest wall motion (KonnoMead
diagram): the technique allows noninvasive measurement
of ventilation and the assessment of the contribution of the
diaphragm and rib cage to tidal volume. The diagram plots
rib cage and abdomen diameters (magnetometers) or perimeters (respiratory inductive plethysmograph). If calibrated,
the technique can measure tidal volume without the need
Advantages
The advantage of this method is that it makes possible the

evaluation of the effectiveness of the muscle, that is, amount
of force generated per stimulation units.
Disadvantages
The disadvantage of this method is the rather complex method

of EMG analysis, which may be made easier by multielectrode
recording technology (33). In addition, this method requires
evaluation of the configuration of the diaphragm via the
KonnoMead method, as well as the velocity of shortening.
This method is still under development, and remains an interesting research tool.
CONCLUSION
This Section of the Statement has considered the mechanical
properties and function of the chest wall, assessed by volume
Figure 6. Abdominal displacement (Vab) gastric pressure (Pga) characteristic during relaxation and contraction of abdominal muscles.
Figure 7. Static progressive contraction of the diaphragm held at three

configurations, as shown in Figure 4 (solid circles). The transdiaphragmatic pressure is plotted against the integrated electrical activity of the
diaphragm, obtained from the crural region via an esophageal electrode and an RC integrator. FRC: The contraction held at FRC configuration. 85% VC: Notice a considerable loss of Pdi per percent Edi
(shorter diaphragm). FRC volume and abdominal muscle contraction
[FRC (ab in)]: Notice the increase in Pdi for any given level of Edi (longest diaphragm). The slopes are an index of the effectiveness of the diaphragm to convert stimulation to pressure. Reprinted by permission
from Reference 28.
587
of a mouthpiece. Initially magnetometers, and later RIP,

have allowed the concept embodied in the KonnoMead
diagram to develop into a clinically useful and widely used
noninvasive measure of ventilation.
Inferring respiratory action from esophageal and gastric pressures (Macklem diagram): the diagram plots esophageal
against gastric pressure and allows breaths that are achieved
mainly by rib cage muscles or diaphragm to be distinguished.
Assessing activation and electromechanical effectiveness of
the diaphragm from EMG (Edi/Pdi diagram): At FRC the
diaphragm can generate much more Pdi for a given Edi
than at high lung volumes. EMG is measured with an esophageal electrode and activity is expressed as a percentage of
EMG at active TLC. At rest, patients with chronic obstructive pulmonary disease have much higher Edi/Pdi ratios than
normal subjects.
The measurements listed above have been widely used in
clinical research and have advanced importantly our understanding of chest wall function. However, relatively little work
has been undertaken in the clinical arena, an important challenge to be met.
References
1. Loring SH, Yoshino K, Kimball WR, Barnas GM. Gravitational and
shear-associated pressure gradients in the abdomen. J Appl Physiol
1994;77:13751382.
2. Decramer M, De Troyer A, Kelly S, Zocchi L, Macklem PT. Regional
differences in abdominal pressure swings in dogs. J Appl Physiol 1984;
57:16821687.
3. Mead J, Yoshino K, Kikuchi Y, Barnas GM, Loring SH. Abdominal
pressure transmission in humans during slow breathing maneuvers. J
Appl Physiol 1990;68:18501853.
4. Agostoni E, Hyatt RE. Static behavior of the respiratory system. In: Macklem PT, Mead J, editors. Handbook of physiology. Vol. 3: The respiratory
system, Part 1. Bethesda, MD: American Physiological Society; 1986. p.
113130.
5. Heaf PJD, Prime FJ. The compliance of the thorax in normal human
subjects. Clin Sci (Colch) 1956;15:319327.
6. Roussos C, Campbell EJM. Respiratory muscle energetics. In: Macklem PT,
Mead J, editors. Handbook of physiology. Vol. 3: The respiratory system,
Part 2. Bethesda, MD: American Physiological Society; 1986. p. 481509.
7. Leblanc P, Bowie DM, Summers E, Jones NL, Killian KJ. Breathlessness and exercise in patients with cardiorespiratory disease. Am Rev
Respir Dis 1986;133:2125.
8. Konno K, Mead J. Measurement of the separate volume changes of rib
cage and abdomen during breathing. J Appl Physiol 1967;22:407422.
9. Smith JC, Mead J. Three degree of freedom description of movement of
the human chest wall. J Appl Physiol 1986;60:928934.
10. McCool FD, Kelly KB, Loring SH, Greaves IA, Mead J. Estimates of
ventilation from body surface measurements in unrestrained subjects.
J Appl Physiol 1986;61:11141119.
11. Mead J, Peterson N, Grimby G, Mead J. Pulmonary ventilation measured from body surface movements. Science 1967;156:13831384.
12. Scanlon PD, Loring SH, Pichurko BM, McCool FD, Slutsky AS, Sarkarati M, Brown R. Respiratory mechanics in acute quadriplegia: lung
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
and chest wall compliance and dimensional changes during respiratory maneuvers. Am Rev Respir Dis 1989;139:615620.
Ringel ER, Loring SH, McFadden ER Jr, Ingram RH Jr. Chest wall configurational changes before and during acute obstructive episodes in
asthma. Am Rev Respir Dis 1983;128:607610.
McCool FD, Loring SH, Mead J. Rib cage distortion during voluntary
and involuntary breathing acts. J Appl Physiol 1985;58:17031712.
Banzett RB, Mahan ST, Garner DM, Brughera A, Loring SH. A simple
and reliable method to calibrate respiratory magnetometers and Respitrace. J Appl Physiol 1995;79:21692176.
Estenne M, De Troyer A. Relationship between respiratory muscle electromyogram and rib cage motion in tetraplegia. Am Rev Respir Dis
1985;132:5359.
Sartene R, Martinot-Lagarde P, Mathieu M, Vincent A, Goldman M,
Durand G. Respiratory cross-sectional areaflux measurements of the
human chest wall. J Appl Physiol 1990;68:16051614.
Cala SJ, Kenyon CM, Ferrigno G, Carnevali P, Pedoti A, Macklem PT,
Rochester DF. Chest wall and lung volume estimation by optical reflectance motion analysis. J Appl Physiol 1996;81:26802689.
Cohen CA, Zagelbaum G, Gross D, Roussos C, Macklem PT. Clinical
manifestations of inspiratory muscle fatigue. Am J Med 1982;73:308316.
Mead J, Loring SH. Analysis of volume displacement and length changes of
the diaphragm during breathing. J Appl Physiol 1982;53: 750755.
Loring SH, Mead J, Griscom NT. Dependence of diaphragmatic length
1985;59:19611970.
Loh L, Goldman M, Newson Davis J. The assessment of diaphragmatic
function. Medicine (Baltimore) 1977;56:165169.
Decramer M, Macklem PT. Pressures developed by the respiratory muscles. In: Roussos C, editor. The thorax. Part B: Applied physiology.
New York: Marcel Dekker; 1995. p. 10991126.
Konno K, Mead J. Static volumepressure characteristics of the rib cage
and abdomen. J Appl Physiol 1968;24:544548.
Goldman MD, Grimby G, Mead J. Mechanical work of breathing derived from rib cage and abdominal VP partitioning. J Appl Physiol
1976;41:752763.
Mead J, Smith JC, Loring SH. Volume displacements of the chest wall
and their mechanical significance. In: Lenfant C, editor. The thorax.
Vol. 85: Lung biology in health and disease. New York: Marcel Dekker; 1995. p. 565586.
Loring SH, Mead J. Action of the diaphragm on the rib cage inferred
from a forcebalance analysis. J Appl Physiol 1982;53:756760.
Grassino A, Goldman MD, Mead J, Sears TA. Mechanics of the human
diaphragm during voluntary contraction: statics. J Appl Physiol 1978;
44:829839.
De Troyer A. The respiratory muscles. In: Cristal R, West J, Weibel E,
Barnes P, editors. The lung: scientific foundations, 2nd ed. Philadelphia, PA: Lippincott-Raven; 1997. p. 12031214.
Lopata M, Evanich MJ, Lorenco R. Quantification of diaphragm EMG
to CO2 rebreathing in humans. J Appl Physiol 1997;43:262270.
Brancatisano A, Engel LA, Loring SH. Lung volume and effectiveness
of inspiratory muscles. J Appl Physiol 1993;74:688694.
Schweitzer TW, Fitzgerald JW, Bowden A, Lynn-Davies P. Spectral analyses
of human diaphragm electromyogram. J Appl Physiol 1979; 46:152165.
Sinderby C, Lindstrom L, Grassino A. Automatic assessment of EMG
quality. J Appl Physiol 1995;79:18031815.
Sinderby C, Beck J, Spahija A, Weinberg J, Grassino A. Voluntary activation of the human diaphragm in health and disease. J Appl Physiol
1998;85:21462158.
7. Imaging Respiratory Muscle Function

This Section of the Statement covers imaging techniques that
pertain to assessment of the function of the muscles of respiration that change the volume of the thorax. It does not consider
modalities for imaging muscles responsible for maintaining
the patency of the upper airway.
TRANSMISSION RADIOGRAPHY
Radiography remains the most used technique for evaluating
the position and movement of the diaphragm. Postero-anterior (P-A) and lateral views of the thorax at total lung capacity
can be supplemented by radiographs at other static lung volumes and movement of the hemidiaphragms as assessed by
fluoroscopy.
Limitations
As shown by three-dimensional reconstruction techniques

using other imaging techniques, the shape of the human diaphragm is complex (13) (Figure 1), so transmission radiography, even with two views, can provide only qualitative estimates
of shape. The silhouette of the diaphragm revealed by transmission radiography represents the most cephalad portion of
the diaphragm apposed to the lung, rather than the chest wall
from one side to the other in P-A views and from dorsal to
ventral in lateral views. The most cephalad portion of the diaphragm, for example, the P-A view, may not lie in the same
dorsalventral plane and, furthermore, may not even represent a contiguous line over the surface of the diaphragm. It
clearly does not represent the curvature of a bundle of muscle
in the diaphragm or necessarily any region of the diaphragm
muscle. Furthermore, attempts to use radiographic changes at
different lung volumes to assess changes in the length of the
diaphragm require assumptions about the relative contribution of the central tendon and muscle. Parallax distortion and
the necessity for exposure to ionizing radiation also cause difficulties with this technology.
Applications
Bearing in mind these important limitations, radiography has

been used to assess diaphragm position and motion in clinical
assessment and to derive estimates of diaphragm length in research studies. In addition, estimating lung volumes from P-A
and lateral radiographs at full inflation may indirectly assist
assessment of the diaphragm.
Position of Hemidiaphragm Domes at Total Lung Capacity
Normal subjects. In 95% of normal adults the level of the

dome of the right hemidiaphragm on postero-anterior radiographs taken standing at full inflation (total lung capacity) is
projected in a plane ranging from the anterior end of the fifth
rib to the sixth anterior interspace; and in only 5% is it at or
below the level of the seventh rib (4). The height of the right
dome tends to be higher in women, in subjects of heavy build,
and in those older than 40 years of age. The plane of the right
diaphragmatic dome tends to be about half an interspace
higher than the left, although in 10% of normal subjects
both are at the same height or the left is higher than the right (5).
Disease. In bilateral diaphragm paralysis both domes are
elevated at total lung capacity (and radiographic lung volume
is
reduced). This change is indistinguishable from volitional
failure
to fully inflate the lungs. In unilateral paralysis, elevaDOI: 10.1164/rccm.AT0702
Internet
www.atsjournals.org
tion
of address:
the paralyzed
dome is obvious. When there is an ac-
quired enlargement of total lung capacity, as in severe emphysema, the domes are lower (the level of the right dome at the
anterior end of the seventh rib or lower) with flattening and a
larger radius of curvature, visible on both P-A and lateral radiographs. If a line is drawn on a P-A radiograph from the vertebrophrenic angle to the costophrenic angle, the maximum
curvature of the dome, assessed at 90 to this line, should normally be at least 1.5 cm. In the most severe disease the domes
may be flat or even inverted with loss of the zone of apposition, allowing visualization on the P-A radiograph of the insertions of the diaphragm into the lower ribs.
Excursion of hemidiaphragm domes during tidal breathing. High-speed cassette changers or video fluoroscopy can
provide dynamic information. In a study of inspiratoryexpiratory radiographs obtained during quiet tidal breathing in the
erect position from 350 subjects, 3080 years of age, and without evidence of respiratory disease, the mean tidal excursions
of the domes of the right and left hemidiaphragms were found
to be 3.3 and 3.5 cm, respectively (6). Tidal diaphragmatic
movement averaged 0.5 cm less in women than in men. Despite similar mean values, unequal movement of the two hemidiaphragmatic domes in an individual subject is common,
most commonly being greater on the right (5).
In bilateral diaphragm paralysis individuals may breathe by
active expiration below relaxation volume followed by passive
inhalation, during which the diaphragm may descend, at least
during early inspiration, leading to erroneous conclusions
about diaphragm function (6, 7). Because relaxation volume
decreases in the supine position, subjects are less likely to be
able to breathe by active expiration below the relaxation volume and passive descent of diaphragm during inspiration is
less likely to occur.
Unilateral diaphragm paralysis is easier to detect because
there is paradoxic motion during tidal inspiration, with ascent
of the paralyzed dome, contrasting with descent of the normal
hemidiaphragm; this contrast can be amplified by the sniff
test, which induces a vigorous, short-lived contraction in the
normal hemidiaphragm. Because of normal variations in movement of the two diaphragms, relative weakness of one hemidiaphragm (such as after cardiothoracic surgery) is difficult to
detect radiographically.
Estimates of Change in Diaphragm Length from Radiographs
at Different Lung Volumes
The length of the silhouette of the diaphragm domes, length of

the diaphragm apposed to the lateral rib cage between the
costophrenic angle and a skeletal landmark corresponding to
its insertion, and the internal diameter of the rib cage can all
be read on a P-A radiograph (8, 9). As discussed above, the silhouette may be formed from the dome in several dorsal-toventral planes, and therefore these measurements do not give
the actual length of the diaphragm in a transverse section. But
it has been argued that changes in these measurements in radiographs obtained at different lung volumes may still give a
reasonable estimate of changes in diaphragm length (8). Similar measurements can also be made with a set of lateral radiographs. To estimate muscle fiber shortening, further assumptions must be made about the uniformity of shortening and the
contribution of the central tendon to the total length. Using
P-A radiographs at several lung volumes, estimated diaphragm
shortening as lung volume increases in normal subjects has
been predicted by a modified piston-and-cylinder model that
Figure 1. Three-dimensional reconstruction, using spiral computed tomography, of diaphragm contour at supine functional residual capacity in a normal subject (A) and a patient with hyperinflation due to
chronic obstructive pulmonary disease (B). The silhouette of the domes
projected on a transmission radiograph may arise from different planes.
Scale in centimeters. Reprinted by permission from Reference 3.
allows for inspiratory expansion of the lower rib cage but neglects change in shape of the domes (9). Radiographic changes
in diaphragm length over the vital capacity have been estimated in patients with chronic obstructive pulmonary disease
(1012). Nonradiographic methods can be used to measure the
internal diameter of the rib cage (calipers or magnetometers,
with chest wall thickness measured by ultrasound) and the
length of the zone of apposition (ultrasound; see below); the
accuracy of estimates of change in the diaphragm length by
nonradiographic methods therefore will be influenced by
whether the dome shape remains reasonably constant (or its
change can be predicted) over the vital capacity (11, 12).
ULTRASOUND
Diaphragm
Ultrasound can be used to image the diaphragm when there is

little or no air intervening between the probe and the muscle.
The normal diaphragm, in common with other skeletal muscles, is poorly echogenic and identification depends on bright
echoes reflected from the attached parietal pleural and peritoneal membranes (13); a further, less bright and irregular echo
may be visible, arising from the layer of connective tissue and
vessels running through the middle of the muscle (14).
Applications
Dome movement. Ultrasound has been used to monitor displacement of the dome of the diaphragm during respiratory
maneuvers, with craniocaudal excursions of the posterior dome
measured either with a transducer on the lateral chest wall
(15) or with a probe placed on the upper abdomen and directed toward the dome (16, 17). Visualization of the right
589
dome is usually better than the left because of the underlying

liver, in contrast to the gastric and intestinal gas adjacent to
the left dome. Diaphragm displacement measured by these
techniques compares favorably with simultaneous fluoroscopic
estimates (18), but the diaphragm is shown as a brightly echogenic
arc with poorly defined borders whose thickness can be as
much as 10 mm; the pleural and peritoneal membranes cannot
be distinguished. In this application a 3.5-MHz vector transducer with considerable penetration but lower resolution is
used in the time-motion mode (M-mode).
Zone of apposition. Ultrasound has also been used to assess the length and thickness of the zone of apposition against
the rib cage; although dynamic changes in length can be assessed by fluoroscopy (19) this entails considerable radiation
exposure. Most of the more recent ultrasound studies have
concentrated on this application; because the costal part of the
diaphragm is relatively close to the skin surface, it is possible
to use a 7.5- or 15-MHz transducer, which has less penetration
but superior resolution. High-resolution M-mode ultrasound
has been used during tidal breathing (20); with two-dimensional (B-mode) ultrasound the two outer echogenic layers of
parietal pleura and peritoneum can be clearly visualized (14)
(Figure 2). In these applications a small ultrasound transducer
is held firmly in a lower intercostal space in the mid- or anterior axillary line (usually on the right side) perpendicular to
the chest wall; if the angle of incidence changes significantly
from perpendicular, resolution is lost. The measurement is
rapid and shows good repeatability in most subjects, but may
be difficult in obese individuals.
With this technique the thickness of the zone of apposition
during relaxation at different lung volumes (14, 21), tidal breathing (20), and static inspiratory efforts (14) has been measured.
Because the diaphragm is identified by its enveloping membranes and there is scattering, ultrasound tends to overestimate diaphragm thickness. However, estimates of the average
thickness of the costal diaphragm in normal untrained subjects
during relaxation at FRC have been in the range of 1.73.3
mm (14, 20, 2224), with larger values in men than women
(22), and lower values in children (24, 25). Healthy newborn
infants have also been studied (26). There appears to be a thicker
diaphragm in individuals with greater inspiratory strength
(27). Unilateral diaphragm paralysis is associated with a thinner costal diaphragm (28), and Duchenne muscular dystrophy
is associated with increased echogenicity (25).
Most measurements have been made in the right anterior
axillary line and little is known about variations in thickness at
different sites in the zone of apposition.
Diaphragm thickening with increasing lung volume should
be inversely proportional to reduction in length; the thickening found as lung volume is increased (14, 21) is compatible
with this relationship, but thickening during tidal breathing
(20, 29) and maximal inspiratory efforts (14) has been greater
than anticipated from the expected shortening. Whether this is
an artifact or is caused by enhanced thickening close to diaphragm insertions into the ribs (as has been shown in dogs) is
not yet known.
Figure 2. B-mode ultrasonogram of the normal diaphragm in the zone of apposition (between the two
arrows) visualized through
a lower right intercostal space.
Scale division on x and y axis
is 10 mm. Reprinted by permission from Reference 14.
590
VOL 166
2002
TABLE 1. RADIOGRAPHY
Name of Test
Information Provided
Diagnostic Purposes
Standard P-A (and lateral)

chest radiograph at TLC
Position of diaphragm domes at TLC

(radiographic estimate of TLC)
Fluoroscopy or rapidradiography technique

during tidal breathing
and sniffs
Movement of right and left

domes: paradoxic motion
exaggerated by brisk
contraction of normal dome
during sniffs
Silhouette of domes, internal
diameter of rib cage, and
estimates of length of zone of
apposition to derive change
in diaphragm length as lung
volume changes
Set of radiographs at
different lung inflation
between RV and TLC
Unilateral/bilateral diaphragm
paralysis
Emphysema
Unilateral or bilateral diaphragm
paralysis
Relationship between diaphragm

length and function, e.g.,
emphysema pre- and post-lung
volume reduction surgery
How to Perform
Standard radiographic techniques
Fluoroscopy or rapid-cassette
change radiography
Standard radiographic techniques

with spirometric control. (Rib
cage diameter can alternatively
be obtained with calipers
or magnetometers and ultrasound; length of zone of
apposition with ultrasound)
Definition of abbreviations: P-A postero-anterior; RV residual volume; TLC total lung capacity.
Potential further applications include the simple assessment

of diaphragm involvement in neuropathy (30) and myopathy, severe acute illness, after cardiothoracic surgery (31), and changes
with rehabilitation and training programs (32). Skeletal muscle
mass is related to force generation. The muscular cross-sectional
area of the diaphragm in the zone of apposition has been estimated from measurements of the internal circumference of the
lower rib cage and measurements of diaphragm thickness (27).
Ultrasound has also been used to detect displacement of
the upper border of the zone of apposition of the costal diaphragm to the rib cage when lung volume changed (8); this
makes use of the loss of diaphragm echo complex, because of

the intervention of aerated lung as lung volume increases. The
length of the zone of apposition can also be estimated by identifying the rib into which the diaphragm is inserted and/or
measuring its upper border at total lung capacity (19). In this
application a longer ultrasound probe is placed craniocaudally
across several ribs and intercostal spaces (again perpendicular
to the chest wall), usually in the anterior axillary line; because
the probe cannot be placed so close to the diaphragm and the
image is lost beneath the ribs, resolution usually is less good
with this technique.
TABLE 2. ULTRASOUND
Name of Test
Information Provided
Diagnostic Purposes
How to Perform
Diaphragm
Dome ultrasound
Movement of right (or left)

dome
Unilateral or bilateral diaphragm

paralysis
Ultrasound probes with

sufficient penetration (3 or
3.5 MHz) placed over
abdomen (A) or over lateral
rib cage (B). M-mode
Zone of apposition ultrasound
Thickness at different lung

volumes, relaxed or
contracted
Detect contraction during tidal

breathing or inspiratory efforts
Effects of pulmonary or
neuromuscular disease, training,
and disuse
Placement of intramuscular
electrodes
High-resolution probe with

less penetration (7.5 MHz)
over intercostal space, usually
in anterior axillary line.
B- or M-mode (C)
Length at different lung

volumes
Estimates of diaphragm length

and swept volume
Linear probe in craniocaudal

plane over lateral rib cage.
B-mode (D)
Both measurements are
usually made on the right
side
Other Respiratory Muscles

Ultrasound of neck, rib cage,
or abdominal muscles
Thickening during
respiratory maneuvers
? Contraction during tidal

breathing or inspiratory efforts
Distance from overlying

skin
Placement of intramuscular
electrodes
High-resolution probe
(7.5 MHz) over appropriate
muscles. B-mode
Other respiratory muscles. Intercostal muscles, parasternal

(33) (and cervical inspiratory) muscles, as well as abdominal
muscles (3436), which are difficult to image by other techniques, can all be imaged with ultrasound, reliance being placed
on echoes from the surrounding fascia. If the relationship between thickness and shortening were known, ultrasound could
be used to assess the pattern of intercostal and abdominal muscle contraction during breathing maneuvers. Ultrasound imaging of these muscles (and the diaphragm) is useful in the accurate placement of intramuscular electrodes to record electrical
activity (28).
VOLUMETRIC IMAGING
Volumetric computed tomography scans (2, 3, 3739) and
magnetic resonance imaging (1, 40, 41) can determine the configuration of the thoracic cavity. Both methods can be used to
determine the detailed shape of the diaphragm (Figure 1), the
rib cage, and ribs in normal (1, 2, 38, 39) and emphysematous
(3, 4244) subjects. Such data have been used with engineering analysis to determine the distribution of tension in the canine diaphragm from measurements of transdiaphragmatic
pressure (37). To date, use of these methods has been limited
to research and few measurements of diaphragm thickness
have been made. Clinical use of both imaging methods is still
limited by the expense of data acquisition and the laborious
data analysis. Computed tomography scans are complicated
by radiation exposure and magnetic resonance imaging is
complicated by prolonged data acquisition times, which must
assemble average data over numerous breaths; both can be applied only in horizontal postures.
NUCLEAR MEDICINE
Both single-photon emission computed tomography (SPECT)
and positron emission tomography (PET) scans provide volumetric information, which can be used to identify the surface
of ventilated or perfused lungs. Because the outer surface of
the lung is apposed to the rib cage and diaphragm, the configuration of the diaphragm and rib cage can also be determined
by these methods. Both techniques offer poor spatial and temporal resolution and there are no published data in which they
have been used to analyze respiratory muscle function.
SUMMARY
Radiographic techniques are widely used to assess respiratory
muscle function (summarized in Table 1). Ultrasound is increasingly used, particularly to assess the diaphragm (Table
2). Volumetric imaging (computed tomography and magnetic
resonance imaging) has limited research application.
References
1. Gauthier AP, Verbanck S, Estenne M, Segebarth C, Macklem PT, Paiva
M. Three-dimensional reconstruction of the in vivo human diaphragm
shape at different lung volumes. J Appl Physiol 1994;76:495506.
2. Pettiaux N, Cassart M, Paiva M, Estenne M. Three dimensional reconstruction of human diaphragm with the use of spiral computed tomography. J Appl Physiol 1997;82:9981002.
3. Cassart M, Pettiaux N, Gevenois PA, Paiva M, Estenne M. Effect of
chronic hyperinflation on diaphragm length and surface area. Am J
4. Lennon EA, Simon G. The height of the diaphragm in the chest radiograph of normal adults. Br J Radiol 1965;38:937943.
5. Young DA, Simon G. Certain movements measured on inspiration
expiration chest radiographs correlated with pulmonary function studies. Clin Radiol 1972;23:3741.
6. Fraser RS, Par JAP, Fraser RG, Par PD. In: Synopsis of diseases of
the chest, 2nd ed. Philadelphia, PA: W.B. Saunders; 1994. p. 98.
591
7. Chen IY, Armstrong JD II. Value of fluoroscopy in patients with suspected bilateral hemidiaphragmatic paralysis. Am J Roentgenol 1993;
160:2931.
8. Loring SH, Mead J, Griscom NT. Dependence of diaphragmatic length
1985;59:19611970.
9. Petroll WM, Knight H, Rochester DF. Effect of lower rib cage expansion and diaphragm shortening on the zone of apposition. J Appl
Physiol 1990;68:484488.
10. Rochester DF, Braun NMT. Determinants of maximal inspiratory pressure in
chronic obstructive pulmonary disease. Am Rev Respir Dis 1985;132:4247.
11. McKenzie DK, Gorman RB, Tolman J, Pride NB, Gandevia SC. Estimation of diaphragm length in patients with severe chronic obstructive
pulmonary disease. Respir Physiol 2000;123:225234.
12. Singh B, Eastwood PR, Finucane KE. Volume displaced by diaphragm
motion in emphysema. J Appl Physiol 2001;91:19131923.
13. Fried AM, Cosgrove DO, Nassiri DK, McReady VR. The diaphragmatic
echo complex: an in vitro study. Invest Radiol 1985;20:6267.
14. Ueki J, DeBruin PF, Pride NB. In vivo assessment of diaphragm contraction by ultrasound in normal subjects. Thorax 1995;50:11571161.
15. Houston JG, Morris AD, Howie CA, Reid JL, McMillan N. Technical
report: quantitative assessment of diaphragmatic movementa reproducible method using ultrasound. Clin Radiol 1992;46:404407.
16. Harris RS, Giovannetti M, Kim BK. Normal ventilatory movement of
the right hemidiaphragm studies by ultrasonography and pneumotachography. Radiology 1983;146:141144.
17. Cohen E, Mier A, Heywood P, Murphy K, Boultbee J, Guz A. Excursionvolume relation of the right hemidiaphragm measured by ultrasonography and respiratory airflow measurements. Thorax 1994;49:885889.
18. Houston JG, Fleet M, Cowan MD, McMillan NC. Comparison of ultrasound with fluoroscopy in the assessment of suspected hemidiaphragmatic movement abnormality. Clin Radiol 1995;50:9598.
19. McKenzie DK, Gandevia SC, Gorman RB, Southon FC. Dynamic
changes in the zone of apposition and diaphragm length during maximal respiratory efforts. Thorax 1994;49:634638.
20. Wait JL, Nahormek PA, Yost WT, Rochester DP. Diaphragmatic thicknesslung volume relationship in vivo. J Appl Physiol 1989;67:15601568.
21. Cohn D, Benditt JO, Eveloff S, McCool FD. Diaphragm thickening during inspiration. J Appl Physiol 1997;83:291296.
22. Taniguchi N, Miyakoda J, Itoh K, Fukui J, Nakamura M, Suzuki O. Ultrasonographic images of the diaphragm and respiratory changes on
their thickness. Jpn J Med Ultrasonics 1991;18:9398.
23. DeBruin PF, Ueki J, Watson A, Pride NB. Size and strength of the respiratory and quadriceps muscles in patients with chronic asthma. Eur
Respir J 1997;10:5964.
24. McCool FD, Benditt JO, Conomos P, Anderson L, Sherman CB, Hoppin FG Jr. Variability of diaphragm structure among healthy individuals. Am J Respir Crit Care Med 1997;155:13231328.
25. DeBruin PF, Ueki J, Bush A, Khan Y, Watson A, Pride NB. Diaphragm
thickness and inspiratory strength in patients with Duchenne muscular dystrophy. Thorax 1997;52:472475.
26. Rehan VK, Nakishima JM, Gutman A, Rubin LP, McCool FD. Effect of
the supine and prone position on diaphragm thickness in healthy term
infants. Arch Dis Child 2000;83:233238.
27. McCool FD, Conomos P, Benditt JO, Cohn D, Sherman CB, Hoppin FG
Jr. Maximal inspiratory pressures and dimensions of the diaphragm.
28. Gottesman E, McCool FD. Ultrasound evaluation of the paralyzed diaphragm. Am J Respir Crit Care Med 1997;155:15701574.
29. Wait JL, Johnson RL. Patterns of shortening and thickening of the human diaphragm. J Appl Physiol 1997;83:11231132.
30. Zifko U, Hartmann M, Girsch W, Zoder G, Rokitansky A, Grisold W,
Lischka A. Diaphragmatic paresis in newborns due to phrenic nerve
injury. Neuropediatrics 1995;26:281284.
31. Dorffner R, Eibenberger K, Youssefzadeh S, Puig S, Liskutin J, Papousek A, Grabenwoger F. The value of sonography in the intensive
care unit for the diagnosis of diaphragmatic paralysis. Rofo Fortschr
Geb Rontgenstr Neuen Bildgeb Verfahr 1998;169:274277.
32. Ayas NT, McCool FD, Gore R, Lieberman SL, Brown R. Prevention of
human diaphragm atrophy with short periods of electrical stimulation.
33. Cala SJ, Kenyon CM, Lee A, Watkin K, Macklem PT, Rochester DF.
Respiratory ultrasonography of human parasternal intercostal muscle
in vivo. Ultrasound Med Biol 1998;24:313326.
34. Strohl KP, Mead J, Banzett RB, Loring SH, Kosch PC. Regional differences in abdominal muscle activity during various maneuvers in humans. J Appl Physiol 1981;51:14711476.
592
35. De Troyer A, Estenne M, Ninane V, Van Gansbeke D, Gorini M. Transversus abdominis muscle function in humans. J Appl Physiol 1990;68:
10101016.
36. Misuri G, Colagrande S, Gorini M, Iandelli I, Mancini M, Duranti R,
Scano G. In vivo ultrasound assessment of respiratory function of abdominal muscles in normal subjects. Eur Respir J 1997;10:28612867.
37. Margulies SS, Lei GT, Farkas GA, Rodarte JR. Finite-element analysis
of stress in the canine diaphragm. J Appl Physiol 1994;76:20702075.
38. Krayer S, Rehder K, Vettermann J, Didier EP, Ritman EL. Position and
motion of the human diaphragm during anesthesia-paralysis. Anesthesiology 1989;70:891898.
39. Whitelaw WA. Shape and size of the human diaphragm in vivo. J Appl
Physiol 1987;62:180186.
40. Gierada DS, Curtin JJ, Erickson SJ, Prost RW, Strandt JA, Goodman
41.
42.
43.
44.
VOL 166
2002
LR. Diaphragmatic motion: fast gradient-recalled-echo MR imaging

in healthy subjects. Radiology 1995;194:879884.
Kanematsu M, Imaeda T, Mochizuki R, Yamawaki Y, Goto H, Sone Y, Doi H.
Dynamic MRI of the diaphragm. J Comput Assist Tomogr 1995;19:6772.
Cassart M, Gevenois PA, Estenne M. Rib cage dimensions in hyperinflated patients with severe chronic obstructive pulmonary disease. Am
Cassart M, Verbandt Y, de Francquen P, Gevenois PA, Estenne M. Diaphragm dimensions after single-lung transplantation for emphysema.
Cassart M, Hamacher J, Verbandt Y, Wildermuth S, Ritscher D, Russi
EW, de Francquen P, Cappello M, Weder W, Estenne M. Effect of lung
volume reduction surgery for emphysema on diaphragm dimensions
and configuration. Am J Respir Crit Care Med 2001;163:11711175.
8. Tests of Upper Airway Function

The upper airway is surrounded by skeletal muscles that exhibit respiratory-related activity (Figures 1 and 2) (14). The
mechanical effects of respiratory-related upper airway muscle
activation are very different from those of the respiratory
pump muscles. Laryngeal muscles regulate the size of the
valvelike glottic aperture. Loss of innervation to the intrinsic
laryngeal muscles can cause stridor leading to respiratory arrest. Pharyngeal muscles stiffen and dilate this potentially collapsible segment of the upper airway. Changes in upper airway muscle activity can alter upper airway caliber and thereby
influence the mechanical effects of respiratory pump muscle
activation. Motor output to upper airway muscles is state dependent. The suppression of upper airway muscle activity during sleep is felt to play an important role in the pharyngeal airway closure during sleep in patients with obstructive sleep
apnea. Electromyography, measurements of upper airway resistance, and upper airway imaging can be used to directly or
indirectly evaluate upper airway muscles. Specific tests are of
use in the clinical evaluation of specific disorders such as extrathoracic airway obstruction and stridor. Evaluation of the
upper airway in the awake patient with obstructive sleep apnea
(OSA) can identify the presence of anatomic abnormalities;
however, upper airway evaluation during wakefulness is of limited value in determining airway function during sleep.
ELECTROMYOGRAPHY
Rationale
Given the complex anatomy of upper airway muscles and

their relatively small size, it is important to verify the correct
placement of the electrodes by having the subject perform voluntary maneuvers associated with activation and/or suppression of the particular muscle. The following maneuvers can be
used to confirm correct electrode placement in some specific
upper airway muscles: genioglossus, activation during tongue
protrusion; pharyngeal constrictors, activation during swallowing; levator palatini, activation on the explosive p during
phonation of repetitive pa sounds; thyroarytenoid, activation
during a prolonged e sound or Valsalva maneuver; posterior
cricoarytenoid, suppression during a prolonged e sound or
Valsalva maneuver and phasic activation on inspiration.
Surface electrodes can be applied on the skin or mucosa
overlying the muscle of interest. Surface electrodes can be applied to the skin in the submental area to record genioglossus
activity and over the cartilaginous portion of the nose to
record alae nasi activity (16, 17). Given the overlapping nature
of upper airway muscles and their relatively small size, surface
electrodes are likely to record electrical activity from more
than one muscle (18). In contrast, intraoral mucosal surface
electrode recordings of genioglossus muscle activity are very
similar to those obtained from hooked-wire electrodes (19).
The electromyogram (EMG) can be analyzed to determine
the presence or absence of respiratory-related activity by comparing it with a signal, such as flow or pressure, that indicates
the phases of respiration (Figure 3). The moving average signal is
used to quantify respiratory-related activity (20). As summarized
by Platt and coworkers (20), the EMG ideally reflects global
activity in a pool of motor neurons of the muscle. The moving
average signal tracks a scaled version of the envelope of the
raw EMG signal. This process is called amplitude demodulation. The moving average is readily interpretable as a continuous indication of the total electrical activity of the muscle, and
digitization of this slow-output waveform requires much less
storage space than the original raw signal. Peak activity is the
maximum activity during a respiratory cycle above electrical
zero. Tonic activity is the minimum activity during a respiratory cycle above electrical zero. Phasic activity during a respiratory cycle is peak minus tonic activity. Measurements are
expressed in arbitrary units, as a percentage of control, or
preferably as a percentage of the maximal activity obtained
during a particular intervention or during the entire study.
The upper airway serves as a common conduit of the respiratory, digestive, and phonatory systems. Whereas respiration requires patency of the upper airway throughout the respiratory
cycle, swallowing and phonation are dependent on upper airway closure. During swallowing, upper airway muscles close the
velopharyngeal sphincter and create a peristaltic-like constriction of the collapsible pharyngeal airway to propel the food bolus into the esophagus. Activation of upper airway muscles during phonation constricts valvelike structures at the velopharynx
and glottis. To maintain airway patency during respiration, activation of upper airway muscles during inspiration dilates and
stiffens the airway. Loss of upper airway muscle activity during
sleep is felt to predispose to pharyngeal airway closure in patients with OSA (Figure 3). Electromyography has been used
extensively in research studies to determine the respiratory-related activation of upper airway muscles and their role in the
pathogenesis of OSA. In particular, information on time domains, coordination among muscles, and relative amounts of
activity (integration) represents the parameters often used (see
ELECTROMYOGRAPHY in Section 3 of this Statement).
The equipment required for EMG recordings of upper airway

muscles is similar to that described in EMG EQUIPMENT in Section 3 of this Statement.
Methodology
Advantages
The techniques used to record the electrical activity of upper

airway muscles are similar to those described in ELECTROMYOGRAPHY in Section 3 of this Statement. Hooked-wire electrodes can be inserted directly into the muscle of interest. A
peroral approach is used to insert the electrode wires into
muscles in the oral cavity and soft palate (511). A transcutaneous approach through the cricothyroid membrane is used to
insert electrode wires into the cricothyroid and thyroarytenoid
muscles (12, 13). A retractable needle catheter advanced
through a fiberoptic scope has been used to insert electrodes
into 10.1164/rccm.AT0802
upper airway muscles that are not easily accessible
DOI:
throughaddress:
a peroral
or transcutaneous approach (14, 15).
Internet
www.atsjournals.org
Electromyographic recordings provide an indirect assessment

of efferent motor output to upper airway muscles. Numerous
research studies have used upper airway electromyography to
determine the role of upper airway muscles in the control of
breathing. Electromyography allows the investigation of individual upper airway muscles. The pharyngeal airway is surrounded by 20 or more skeletal muscles. Electromyographic
recordings have been obtained from only a minority of these
muscles because of their relative inaccessibility. The genioglossus is the most frequently recorded upper airway muscle because of its easy accessibility. However, significant differences
exist in the electromyographic activation of upper airway mus-
Equipment
594
VOL 166
2002
Figure 2. Intrinsic laryngeal

muscles attached to the
arytenoid cartilage help regulate the size of the glottic
aperture.
cles, and genioglossus activity may not always be representative of motor output to other upper airway or even pharyngeal
muscles. For example, during quiet breathing, the pharyngeal
constrictors rarely exhibit respiratory-related activity, whereas
the posterior cricoarytenoid, a vocal cord abductor, uniformly
exhibits phasic inspiratory activity, and the thyroarytenoid, a
vocal cord adductor, exhibits phasic expiratory activity (7, 13, 21).
wires. Reproducibility for the wire electrodes was 0.62, and

this dropped to 0.22 when the electrodes were reapplied. Jonsson and Komi (28) also found poor reproducibility in the recorded EMG signal among four pairs of wire electrodes implanted in the same brachioradialis muscle.
Comparative studies of upper airway function during respiration have frequently used the genioglossus as a representative upper airway muscle (11, 22, 29). However, the genioglossus is fixed at only one end, with many degrees of freedom of
movement. During protrusions against a balloon transducer to
measure force, control of tongue length and orientation cannot be assured. Likewise, it cannot be assumed that the contraction during voluntary maneuvers is isometric or even consistent from trial to trial.
Disadvantages
Applications
Upper airway EMGs are technically difficult given the complex

anatomy of these relatively small muscles combined with their
often conflicting respiratory and nonrespiratory functions.
Electromyography of a specific upper airway muscle does not
reveal the mechanical effects of this electrical activation or how
the electrical activation of a particular upper airway muscle interacts with the simultaneous electrical activation of other upper airway muscles to influence upper airway function. For example, during quiet breathing in wakefulness, the posterior
cricoarytenoid, a vocal cord abductor, is tonically active
throughout the respiratory cycle and exhibits phasic activity on
inspiration. In contrast, the thyroarytenoid, a vocal cord adductor, is also tonically active throughout the respiratory cycle but
exhibits phasic activation on expiration. Electromyography
does not reveal the net effect of these simultaneously active
counteracting muscles on vocal cord position. This example
further emphasizes that electromyography of one specific upper airway muscle is not necessarily representative of others.
Another general weakness of electromyography is the inability to calibrate the recordings. As a result it is very difficult
to compare EMG results from a specific muscle between days
or between subjects (22). Some studies have attempted to circumvent this problem by expressing the data as a percentage
of maximum activation (10, 11). Such comparisons, however, are
problematic. Assuming that filtering and gains are fixed, EMG
recording is affected by factors such as electrode site, distance
between electrodes, and skin resistance (surface electrode) (22).
Normalization techniques are commonly used in muscle research to make comparisons across nonrespiratory muscle
groups (2226). However, in this regard, important differences
exist between limb muscles and upper airway muscles. Limb
muscles are attached to bone at both ends and as such, muscle
length and orientation can be fixed so that the contractions
are isometric. In addition, EMG studies of limb muscles generally normalize the data to known EMGforce relationships.
Using wire electrodes to record from the biceps brachii, Komi
and Biskirk (27) examined reproducibility within days with
the wires unchanged and across days with reapplication of the
The clinical indications for electromyography of upper airway

muscles are limited, and in general the technique should be
reserved for research studies. Submental surface electrodes
are used to record muscle activity as a qualitative signal during
polysomnograms to stage sleep. Vocal cord EMG recordings
are routinely performed in many centers to assess vocal cord
paralysis. By contrast, in general, EMG of the upper airway
muscles in research studies is the subject of a more quantitative evaluation and measurement.
Figure 1. Some muscles that influence the pharyngeal airway. Contraction of muscles attaching to the hyoid bone can displace the anterior pharyngeal wall ventrally, helping to increase airway size. Muscles
attached to the soft palate help regulate the route of breathing.
UPPER AIRWAY RESISTANCE

Rationale
Changes in upper airway muscle activity can influence upper

airway caliber (16, 30). Therefore, measurements of upper airway resistance can be used to indirectly assess changes in upper airway muscle activity. However, upper airway resistance
is also influenced by changes in mucosal vasculature, body position (e.g., supine versus prone), head position (e.g., extended
versus flexed), lung volume, and application of continuous
positive airway pressure (31, 32). Upper airway resistance also
has a direct but nonlinear relationship with flow.
Methodology
Resistance is calculated by measuring the pressure drop across

the airway at a given flow. The upper airway includes the larynx,
pharynx, and nasal and oral cavities. Upper airway resistance can
be measured during oral or nasal breathing. Determination of total upper airway resistance requires measurement of the pressure drop between the subglottic airway and the nares or mouth.
Resistance across the nasal, pharyngeal, or laryngeal segments of
the upper airway can be determined by measuring the pressure
drop across the particular segment of interest (33). Measurement
of subglottic pressure requires the percutaneous or transnasal
placement of a catheter attached to a pressure transducer into
the extrathoracic trachea (34). Because of the technical difficulty
of both methods, most studies of upper airway resistance exclude
the laryngeal airway from their measurements.
595
Figure 3. Electromyogram (EMG) of the genioglossus (GG) muscle during two obstructive apneas
(OSAs) in nonrapid eye movement sleep. Airway
closure (horizontal lines) is associated with a decrease in GG activation. GG activation progressively increases during the obstructive, apneas but
airway reopening, as evidenced by the resumption
of tracheal breathing sounds, does not occur until there is a large burst of GG activity. The time
constant of the oxygen saturation signal delays
the appearance of the oxygen saturation nadir
until after airway reopening. EEG electroencephalogram; SaO2 arterial oxygen saturation.
Supraglottic airway resistance is the pressure drop from the

level of the epiglottis to the airway opening (mouth or nose) at
a given flow (Figure 4). A face mask connected to a pneumotachograph is attached via an airtight seal to the individual to
measure flow. Mask pressure, that is, pressure at the nares or
mouth, is measured from a port in the mask. The intralumenal
airway pressure measurement requires an invasive procedure.
After topical anesthesia of one nasal passage, a catheter attached to a pressure transducer is advanced to the level of the
epiglottis and secured at the nose. To eliminate the influence
of kinetic energy, the pressure is measured from a side port
near the sealed tip of the catheter. Because changes in the position of the catheter can influence the measurements, studies
should be designed so that the control and experimental recordings are obtained without repositioning or removing the
catheter. The catheter can be filled with air or water. An airfilled catheter has the potential to become occluded with fluid.
Once in position, pressure from the water-filled catheter at zero
flow must be obtained to determine the contribution of hydrostatic pressure to the measurements. Pressure transducer-tipped
catheters have also been used to measure pharyngeal pressure.
However, these catheters work best in a completely dry or
fluid environment such as the intravascular space and esophagus.
The repetitive wetting and drying of the transducer surface in the
pharynx frequently results in an unacceptable drift of the signal.
Upper airway resistance is calculated by dividing the pressure
drop across the airway at a particular flow by that flow. The units
of measurement are cm H2O L1 second1. It is important
to state whether the measurements are obtained during inspiration or expiration as, even at the same flow, upper airway
resistance during inspiration is not the same as upper airway
resistance during expiration. Measurements on different breaths
should also be obtained from the same portion of inspiration
or expiration as the pressureflow relationships during each of
the two respiratory phases may show hysteresis.
An alternative method used to quantify upper airway resistance is to describe the entire pressureflow curve on inspiration or expiration by fitting the pressure and flow data to a
second-order equation. Numerous equations have been used,
including the Rohrer equation: P K1V K2V2 (Equation 1),
where P is pressure, V is flow, and K1 and K2 are constants.
The presence of two constants in this equation makes statistical comparisons difficult and, therefore, another frequently
used equation is: P KV2 (Equation 2). Changes in upper airway resistance can be assessed indirectly by measuring total
pulmonary resistance with the passage of an esophageal balloon. This method is valid if resistance across the lower airways remains constant. Therefore, these measurements must
be obtained at a given lung volume.
The measurement of upper airway resistance requires two
pressure transducers to measure the pressure drop across the
airway and a pneumotachograph connected to a differential
pressure transducer for the measurement of flow. Pressure is
calibrated in cm H2O and flow in L second1. As the measurements are obtained under dynamic conditions, the three
signals should remain in phase up to a frequency of 10 Hz.
Advantages
When properly controlled, measurements of upper airway resistance can provide a global assessment of the effect of changes
in upper airway muscle activity on upper airway function.
Figure 4. Changes in upper airway resistance in the breaths preceding

an obstructive apnea in nonrapid eye movement sleep. The supraglottic pressureflow plot shows a progressive increase in inspiratory resistance, i.e., narrowing of the pharyngeal airway, on the four breaths
preceding complete airway closure on inspiration (breath 5).
596
Disadvantages
FIBEROPTIC IMAGING
Upper airway resistance is only an indirect measurement of upper airway muscle activity. Measurements must be carefully controlled as upper airway resistance can be affected by many other
factors (see previous sections). Measurement of upper airway resistance requires an invasive procedure that is of low morbidity.
Rationale
Applications
The determination of upper airway resistance is a clinical research technique that has been used to assess the effect of changes
in chemical drive and sleepwakefulness states on upper airway size and muscle activity. Nasal rhinometry, the measurement of nasal airway resistance, is used by ear, nose, and throat
physicians to assess the patency of the nasal passages.
INDIRECT LARYNGOSCOPY
Rationale
Visualization of the glottis by indirect laryngoscopy can confirm

vocal cord paralysis resulting from the loss of intrinsic laryngeal
muscle innervation. All the intrinsic laryngeal muscles, except the
cricothyroid, are innervated by the recurrent laryngeal nerve.
Loss of motor innervation from the recurrent laryngeal nerve, as
may occur secondary to a mediastinal tumor or as a complication
of thyroidectomy, results in vocal cord paralysis and hoarseness.
With bilateral vocal cord paralysis, the cords are positioned close
to the midline and show no abduction during inspiration. Audible stridor is almost invariably present. Stridor also occurs during
the laryngospasm and/or laryngoedema that can be encountered
immediately after removal of an endotracheal tube.
Methodology
Indirect laryngoscopy is performed in the awake patient by advancing a small angled mirror through the open mouth to the
soft palate while protruding the subjects tongue. Illumination of
the angled reflecting surface allows visualization of the glottic aperture through the oral cavity. Topical anesthesia of the soft palate can help minimize the gag reflex. As the cords are being visualized, the patient is instructed to vocalize an e sound that
normally adducts the vocal cords. Failure of one cord to adduct
indicates ipsilateral vocal cord paralysis. In the case of bilateral
vocal cord paralysis, both cords are close to the midline and show
no abduction during inspiration and may even show paradoxical
movement with respiration due to the effects of changes in intralumenal pressure on inspiration and expiration. Direct laryngoscopy is performed under general anesthesia. With the advent
of fiberoptic scopes, this latter technique is rarely indicated in
adults to assess the presence or absence of vocal cord paralysis.
Advantages
Indirect laryngoscopy is an easy technique that requires little

equipment. It is performed in awake patients. It provides a direct visual assessment of vocal cord movement and is the technique of choice for confirmation of vocal cord paralysis if a fiberoptic instrument is not available.
VOL 166
2002
Fiberoptic imaging can be performed in place of indirect

laryngoscopy for detection of vocal cord paralysis. In clinical
research, fiberoptic imaging is used to evaluate the mechanical
effects of laryngeal and pharyngeal muscle activation on upper airway size. Fiberoptic imaging has also been used in research laboratories to assess upper airway mechanics in the
absence of upper airway muscle activity (35, 36).
Methodology
Fiberoptic imaging is an invasive procedure that provides direct visualization of the upper airway. However, it is easily
performed in the awake adult patient and is associated with
very little discomfort. After topical anesthesia of one nasal
passage, the fiberoptic scope is advanced through the nares
and along the floor of the nasal passage into the pharyngeal
airway. Depending on the outer diameter of the scope, prior
application of a nasal decongestant may be of benefit. The
glottic aperture is easily visualized from the hypopharynx. No
attempt should be made to touch or pass through the unanesthetized glottis during this visual examination.
Equipment
Application as a diagnostic technique requires a fiberoptic nasopharyngoscope, optimally with an outer diameter of 4 mm or
less, attached to a light source. Physicians performing the procedure should be specially trained in this technique.
When using fiberoptic imaging as a research technique to
quantify changes in upper airway dimensions, it is necessary to
have a camera attached to the fiberoptic scope to obtain videotape recordings. Individual frames of the videotape can
then be analyzed offline, ideally with a personal computer
frame grabber and digitizing software. Methods have been described to calibrate the measurements in metric units (36, 37).
Advantages
Compared with indirect laryngoscopy, fiberoptic examination

of the upper airway allows a more thorough examination of
the upper airway. If a video camera is available, the physical
findings can be documented on videotape. This technique is
used for clinical diagnostic purposes.
Disadvantages
The fiberoptic equipment is expensive. Standardized techniques should be instituted to sterilize the scope between procedures. Passage of the fiberoptic scope may be associated
with a vasovagal reaction associated with hypotension and loss
of consciousness. Adequate supportive care should be available in the event of this complication.
Use of fiberoptic imaging of the upper airway in clinical research is laborious even with computer-assisted analysis. In
addition to difficulties with calibration of measurements in
metric units, the lack of depth perception on the videotape images makes it particularly difficult to detect the edge of the
pharyngeal airway that should be measured. Movement and
clouding of the scope are other technical difficulties frequently encountered during research studies.
Disadvantages
As a visual assessment of intrinsic laryngeal muscle activity, indirect laryngoscopy is limited to determining the presence or absence of motor innervation. Bilateral but asymmetrical movement
of the vocal cords is difficult to quantify. This finding may suggest
vocal cord paresis but can also arise from other conditions.
COMPUTED TOMOGRAPHIC SCANNING

Rationale
Computed tomographic (CT) scanning allows imaging of the

upper airway size and anatomy (38, 39). The technique may
597
also demonstrate related structural abnormalities, which may

impinge on the upper airway.
Equipment
Routine CT scanners are used (see Section 7 of this Statement). Three-dimensional reconstruction may be helpful.
Advantages
Computed tomography is widely available, it is noninvasive,

and it provides images of the whole upper airway.
Disadvantages
The subject receives radiation exposure. Subjects are usually

studied awake, whereas the object of investigation is often the
site of upper airway narrowing during sleep.
Advantages
Acoustic reflection is noninvasive, nonirradiating, inexpensive, compact, and can be performed rapidly. It can be used in
large numbers of subjects in a variety of different body postures (46).
Disadvantages
Acoustic reflection cannot be used to visualize the retropalatal airway and cannot be used during sleep.
Applications
Acoustic reflection is applicable to population or physiological studies of upper airway size. It is of little value to individual patients.
Applications
FLOWVOLUME LOOPS
Computed tomography is of limited value in determining the

site of obstruction of the upper airway, as this must be done
during sleep; the site of maximal narrowing during wakefulness may not be the site of occlusion during sleep. Faster acquisition CT scanners hold the promise of providing not just
structural images of the upper airway but also the ability to assess upper airway function.
Rationale
MAGNETIC RESONANCE IMAGING

Rationale
Magnetic resonance imaging (MRI) scanning allows imaging

of the upper airway size and anatomy (40). The technique may
also demonstrate related structural abnormalities that may
impinge on the upper airway (4143).
Flowvolume loops have the potential to be used to detect upper airway obstruction during wakefulness as a predictor of
airway occlusion during sleep.
Equipment
Recording of flowvolume loops requires a body plethysmograph, rapid response spirometer, or pneumotachograph (see
Advantages
Flowvolume loops can be measured quickly. The equipment

is widely available and measurements can be made during
wakefulness.
Equipment
Disadvantages
Routine MRI scanners are used (see Section 7 of this Statement).
Flowvolume loops cannot usefully distinguish patients with

sleep apnea from others (47, 48).
Advantages
MRI is widely available. It is noninvasive and provides images

of the whole upper airway. It is not ionizing and can identify
adipose deposits around the airway.
Disadvantages
Slow scanning times on many scanners mean that several

breath cycles are averaged. Subjects are usually studied awake,
whereas it is the site of upper airway narrowing during sleep
that is often the object of investigation.
Applications
Magnetic resonance imaging is of limited value in determining

the site of obstruction of the upper airway, as this must be done
during sleep; the site of maximal narrowing during wakefulness
may not be the site of occlusion during sleep. Like CT, MRI
holds the promise of providing not just structural images of the
upper airway but also the ability to assess upper airway function.
ACOUSTIC REFLECTION
Rationale
Acoustic reflection allows quantification of upper airway caliber

by noninvasive means. The subject breathes through a mouthpiece attached to a microphone that emits sound waves. The reflected impulse is analyzed to provide a plot of the cross-section of
the upper airway as a function of distance from the mouth (44, 45).
The system requires a loudspeaker, microphone, wave tube,
mouthpiece shutter system, and personal computer. Commercial systems are available.
Applications
The only clinical role is as a quick method of assessing whether a

gross mass lesion is present around the upper airway. It is not of
routine value.
POLYSOMNOGRAPHY
Rationale
Polysomnography is the study of several physiological variables during sleep, including the recording of sleep itself. The
precise items recorded will vary but usually include electroencephalography, electro-oculography, electromyography, respiratory pattern, snoring, oxygen saturation, transcutaneous carbon dioxide tension, electrocardiography, and body posture
(4951). This allows the recognition of events that occur during sleep and offers the advantage over other techniques that
sleep can be identified with certainty.
In patients with marked respiratory muscle weakness, detection of respiratory movement during obstructive events from
external sensors may be very difficult or impossible. Indeed,
true obstructive events may be scored even by skilled observers
as central. There are three ways around this problem. First,
esophageal pressure may be recorded during polysomnography in all such patients (52, 53). Second, flattening of the flow
time curve during these events in sleeping subjects may be
helpful (54). Third, some centers have adopted the approach
that patients with multiple central apneas should be treated
with continuous positive airway pressure to determine whether
these events respond to continuous positive airway pressure.
598
Equipment
Applications
Computerized systems are generally used, which allows the recording of the all-night data on to optical or compact disk formats. Detection of fluctuation in nasal pressure during inspiration and expiration reflects changes in inspiratory and expiratory
airflow and therefore is a promising method for the detection
of hypopneas (50, 55). Studies have shown that nasal pressure
is more sensitive than thermal sensors for detecting hypopneas and that the square root of the pressure signal improves
the estimate of airflow (56, 57).
At present, applications are limited to research.
Advantages
Polysomnography allows identification of sleep. It can detect

coincidence of different events, such as respiratory change and
sleep disturbance. It is possible to identify the extent of decline in
arterial oxygen saturation (SaO2) during different sleep stages.
The sequelae of hypoventilation due to respiratory muscle insufficiency during sleep can be identified.
Disadvantages
Polysomnography is expensive relative to limited sleep study

equipment, both in terms of capital and staff cost. Expert technical staff are needed to perform and analyze the studies.
Applications
Polysomnography is a valuable method of identifying sleeprelated hypoventilation in patients with respiratory muscle
problems and whether they have related sleep apneahypopnea syndrome as well. It is useful in some patients to monitor
progress on treatment, such as noninvasive positive pressure
ventilation.
MUSCLE BIOPSY
Rationale
Differences in upper airway muscle fiber composition have

been reported in sleep apneahypopnea syndrome patients
(58, 59).
VOL 166
2002
SITE OF PHARYNGEAL AIRWAY CLOSURE DURING SLEEP

Rationale
Loss of upper airway muscle activity during sleep is thought to

have an important role in the pathogenesis of obstructive
sleep apnea. However, because of the differences in upper airway muscle activity between wakefulness and sleep, studies of
the upper airway during wakefulness are not capable of predicting the site of upper airway closure during sleep. Airway
closure during sleep in patients with sleep apnea can occur at
different sites in the pharyngeal airway (36, 60, 61). The primary site of closure is most frequently located in the velopharynx but can also occur in the oro- and/or hypopharynx. Determination of the site of airway closure is important, as studies
indicate that it determines the outcome of uvulopalatopharyngoplasty (62, 63). Tests that determine the site of upper airway
closure during sleep cannot be used to assess upper airway
muscle function, as airway closure during sleep is also dependent on other factors, including the transmission of subatmospheric intrathoracic pressure into the pharyngeal airway and
closing pressure (i.e., the pressure at which the pharyngeal airway closes in the absence of upper airway muscle activity and
respiratory effort) (15, 27). Factors that narrow the pharyngeal
airway raise closing pressure. Normally, closing pressure is
subatmospheric. In patients with OSA, closing pressure is usually
positive, that is, a positive intralumenal pressure is required
during sleep to maintain pharyngeal airway patency. The site of
airway closure can be determined during sleep with the placement of catheters attached to pressure transducers in the esophagus and at different levels in the pharyngeal airway (57). The site
of pharyngeal airway closure during sleep can also be determined
by imaging the upper airway with a fiberoptic scope or by fast
acquisition MRI or computed axial tomography scans (36, 64).
CONCLUSION
A force transducer and EMG (see above) are required.
Structurally, the upper airway consists of several valves (nares,

mouth, soft palate, collapsible pharynx, epiglottis, and glottis)
that are regulated in part by upper airway muscles. These
valves allow airway closure during deglutition and phonation
and regulate airway patency during respiration.
Two of the most important upper airway disorders affecting respiration are laryngeal narrowing and OSA. Laryngeal
narrowing can result from laryngospasm and vocal cord paralysis. Tests of upper airway function that detect laryngeal narrowing and dysfunction include indirect laryngoscopy, fiberoptic imaging, electromyography, and flowvolume loops.
The upper airway closure in the vast majority of patients with
OSA occurs in the pharynx with suppression of upper airway
muscle activity during sleep.
Although polysomnography is not a specific test of upper
airway function, it is the accepted method to diagnose this repetitive closure of the pharyngeal airway during sleep. A plethora of techniques have been developed to assess respiratoryrelated upper airway structure and function, ranging from
measurements of upper airway resistance to MRI. However,
in general, these do not have a role in the clinical assessment
of upper airway function and should be reserved for the research laboratory.
Disadvantages
References
Technique
Muscle biopsy is not generally performed as a diagnostic test,

and is reserved for research examination of resected specimens.
Advantages
Muscle biopsy has no clinical applications.

Disadvantages
It is invasive and painful.

Applications
At present, applications are limited to research.
STRENGTH, FATIGUE, AND ENDURANCE OF UPPER

AIRWAY MUSCLES
Rationale
It is not known whether strength, fatigue, and endurance of

upper airway muscles are factors in sleep-related upper airway
obstruction (54).
Equipment
The tongue is the only upper airway muscle for which force
can be measured.
1. Van Lunteren E, Strohl KP. The muscles of the upper airways. Clin
Chest Med 1986;7:171188.

2. Bartlett D Jr. Respiratory functions of the larynx. Physiol Rev 1989;
69:3357.
3. Bartlett D Jr. Upper airway motor systems. In: Fishman AP, Cherniack
NS, Widdicombe JG, Geiger SR, editors. Handbook of physiology:
the respiratory system, Section 3, Vol. II. Bethesda, MD: American
Physiological Society; 1986. p. 223245.
4. Mathew OP, Sant-Ambrogio G, editors. Respiratory function of the upper airway. In: Lenfant C, editor. Lung biology in health and disease,
Vol. 35. New York: Marcel Dekker; 1988.
5. Basmajian JV, Stecko G. A new bipolar electrode for electromyography.
6. Sauerland EK, Harper RM. The human tongue during sleep: electromyographic activity of the genioglossus muscle. Exp Neurol 1976;51:160170.
7. Launois SH, Tsui J, Weiss JW. Respiratory function of velopharyngeal
constrictor muscles during wakefulness in normal adults. J Appl Physiol 1997;82:584591.
8. Tangel DJ, Mezzanotte WS, Sandberg EJ, White DP. Influences of
NREM sleep on the activity of tonic vs. inspiratory phasic muscles in
normal men. J Appl Physiol 1992;73:10581066.
9. Tangel DJ, Mezzanotte WS, White DP. Influence of sleep on tensor palatini EMG and upper airway resistance in normal men. J Appl Physiol 1991;70:25742581.
10. Mezzanotte WS, Tangel DJ, White DP. Influence of sleep onset on upper-airway muscle activity in apnea patients versus normal controls.
11. Mezzanotte WS, Tangel DJ, White DP. Waking genioglossal electromyogram in sleep apnea patients versus normal controls (a neuromuscular
compensatory mechanism). J Clin Invest 1992;89:15711579.
12. Wheatley JR, Brancatisano A, Engel LA. Respiratory-related activity of
cricothyroid muscle in awake normal humans. J Appl Physiol 1991;70:
22262232.
13. Kuna ST, Insalaco G, Woodson GE. Thyroarytenoid muscle activity
during wakefulness and sleep in normal adults. J Appl Physiol 1988;
65:13321339.
14. Kuna ST, Smickley JS, Insalaco G. Posterior cricoarytenoid muscle activity during wakefulness and sleep in normal adults. J Appl Physiol
1990;68:17461754.
15. Kuna ST, Insalaco G, Villeponteaux RD. Arytenoideus muscle activity
in normal adult humans during wakefulness and sleep. J Appl Physiol
1991;70:16551664.
16. Strohl KP, OCain CF, Slutsky AS. Alae nasi activation and nasal resistance in health subjects. J Appl Physiol 1982;52:14321437.
17. Suratt PM, McTier R, Wilhoit SC. Alae nasi electromyographic activity
and timing in obstructive sleep apnea. J Appl Physiol 1985;58:12521256.
18. Insalaco G, Sant-Ambrogio G, Sant-Ambrogio FB, Kuna ST, Mathew
OP. Specificity of esophageal electrode recordings of posterior cricoarytenoid muscle activity. J Appl Physiol 1989;66:15011505.
19. Doble EA, Leiter JC, Knuth SL, Daubenspeck JA, Bartlett D Jr. A noninvasive intraoral electromyographic electrode for genioglossus muscle. J Appl Physiol 1985;58:13781382.
20. Platt RS, Hajduk EA, Hulliger M, Easton PA. A modified Bessel filter
for amplitude demodulation of respiratory electromyograms. J Appl
Physiol 1998;84:378388.
21. Kuna ST, Smickley JS, Vanoye CR. Respiratory-related pharyngeal constrictor muscle activity in normal human adults. Am J Respir Crit Care
Med 1997;155:19911999.
22. Suratt PM, McTier RF, Wilhoit SC. Upper airway muscle activation is
augmented in patients with obstructive sleep apnea compared with
that in normal subjects. Am Rev Respir Dis 1988;137:889894.
23. Basmajian JV, De Luca CJ. Muscles alive: their functions revealed by
electromyography. Baltimore, MD: Williams & Wilkins; 1985.
24. Lawrence JH, De Luca CJ. Myoelectric signal versus force relationship
in different human muscles. J Appl Physiol 1983;54:16521659.
25. Lippold OCJ. The relation between integrated action potentials in a human muscle and its isometric tension. J Physiol 1952;117:492499.
26. Zuniga EN, Truong XT, Simons DG. Effects of skin in electrode position on averaged electromyographic potentials. Arch Phys Med 1970;
51:264272.
27. Komi PV, Biskirk ER. Reproducibility of electromyographic measurements with inserted wire electrodes and surface electrodes. Electromyography 1970;10:357367.
28. Jonsson B, Komi PV. Reproducibility problems when using wire electrodes in electromyographic kinesiology. In: New developments in
electromyography and clinical neurophysiology. Basel, Switzerland: S.
Karger; 1973. p. 540.
29. Leiter JC, Knuth SL, Bartlett D Jr. The effects of sleep deprivation on activity of the genioglossus muscle. Am Rev Respir Dis 1985;132:12421245.
599
30. Anch AM, Remmers JE, Bunce H III. Supraglottic airway resistance in
normal subjects and patients with occlusive sleep apnea. J Appl Physiol 1982;53:11581163.
31. Wasicko MJ, Hutt DA, Parisi RA, Neubauer JA, Mezrich R, Edelman
NH. The role of vascular tone in the control of upper airway collapsibility. Am Rev Respir Dis 1990;141:15691577.
32. White DP, Lombard RM, Cadieux RJ, Zwillich CW. Pharyngeal resistance in normal humans: influence of gender, age, and obesity. J Appl
Physiol 1985;58:365371.
33. Hudgel DW, Martin RJ, Johnson B, Hill P. Mechanics of the respiratory
system and breathing pattern during sleep in normal humans. J Appl
Physiol 1984;56:133137.
34. Kuna ST, Vanoye CR, Griffin JR, Updegrove JD. Effect of hypercapnia
on laryngeal airway resistance in normal adult humans. J Appl Physiol
1994;77:27972803.
35. Isono S, Morrison DL, Launois SH, Feroah TR, Whitelaw WA, Remmers JE. Static mechanics of the velopharynx of patients with obstructive sleep apnea. J Appl Physiol 1993;75:148154.
36. Morrison DL, Launois SH, Isono S, Feroah TR, Whitelaw WA, Remmers JE. Pharyngeal narrowing and closing pressures in patients with
obstructive sleep apnea. Am Rev Respir Dis 1993;148:606611.
37. Kuna ST, Vanoye CR. Laryngeal response during forced vital capacity
maneuvers in normal adult humans. Am J Respir Crit Care Med 1994;
150:729734.
38. Haponik EF, Smith PL, Bohlman ME, Allen RP, Goldman SM, Bleecker
ER. Computerized tomography in obstructive sleep apnea: correlation
of airway size with physiology during sleep and wakefulness. Am Rev
Respir Dis 1983;127:221226.
39. Shepard JW, Thawley SE. Evaluation of the upper airway by computerized tomography in patients undergoing uvulopalatopharyngoplasty
for obstructive sleep apnea. Am Rev Respir Dis 1989;140:711716.
40. Rodenstein DO, Dooms G, Thomas Y, Liistro G, Stanescu DC, Culee C,
Aubert-Tulkens G. Pharyngeal shape and dimension in healthy subjects, snorers and patients with obstructive sleep apnoea. Thorax 1990;
45:722727.
41. Horner RL, Mohiaddin RH, Lowell DG, Shea SA, Burman ED, Longmore DB, Guz A. Sites and sizes of fat deposits around the pharynx in
obese patients with obstructive sleep apnoea and weight matched controls. Eur Respir J 1989;2:613622.
42. Shelton KE, Gay SB, Hollowell DE, Woodson H, Suratt PM. Mandibular enclosure of upper airway and weight in obstructive sleep apnea.
43. Shelton KE, Woodson H, Gay S, Suratt PM. Pharyngeal fat in obstructive sleep apnea. Am Rev Respir Dis 1993;148:462466
44. Fredberg JJ, Wohl MB, Glass GM, Dorkin HL. Airway area by acoustic
reflections measured at the mouth. J Appl Physiol 1980;48:749758.
45. Bradley TD, Brown IG, Grossman RF, Zarnel N, Martinez D, Phillipson
EA, Hoffstein V. Pharyngeal size in snorers, nonsnorers, and patients
with obstructive sleep apnea. N Engl J Med 1986;315:13271331.
46. Martin SE, Marshall I, Douglas NJ. The effect of posture on airway calibre in patients with the sleep apnea/hypopnea syndrome. Am J Respir
Crit Care Med 1995;152:721724.
47. Riley R, Guilleminault C, Herran J, Powell N. Cephalometric analyses
and flow volume loops in obstructive sleep apnea patients. Sleep 1983;
6:303311.
48. Krieger J, Weitzenblum E, Vandevenne A, Stierle J, Kurtz D. Flow volume curve abnormalities and obstructive sleep apnea syndrome. Chest
1985;87:163167.
49. Douglas NJ, Thomas S, Jan MA. Clinical value of polysomnography.
Lancet 1992;339:347350.
50. American Academy of Sleep Medicine Task Force. Sleep-related breathing disorders in adults: recommendations for syndrome definitions and
measurement techniques in clinical research. Sleep 1999;22:667689.
51. American Sleep Disorders Association. Practice parameters for the indications for polysomnography and related procedures. American Sleep
Disorders Association Report. Sleep 1997;20:406422.
52. Tvinnereim M, Mataeika S, Cole P, Height J, Hoffstein V. Diagnosis of
obstructive sleep apnea using a portable transducer catheter. Am J
53. Guilleminault C, Stoohs R, Clerk A, Cetel M, Maistros P. A cause of excessive daytime sleepiness: the upper airways resistance syndrome.
Chest 1993;104:781787.
54. Mortimore IL, Douglas NJ. Genioglossal strength and fatiguability: relationship to apnea/hypopnea index. Am J Respir Crit Care Med 1996;
153:A532.
55. Condos R, Norman RG, Krishnasamy I, Petuzzi N, Goldring RM, Rapoport DM. Flow limitation as a noninvasive assessment of residual upper-
600
56.
57.
58.
59.
60.
airway resistance during continuous positive airway pressure therapy of

obstructive sleep apnea. Am J Respir Crit Care Med 1994;150:475480.
Hosselet JJ, Norman RG, Ayappa I, Rapoport DM. Detection of flow
limitation with a nasal cannula/pressure transducer system. Am J Respir
Crit Care Med 1998;157:14611467.
Norman RG, Ahmed MM, Walsleben JA, Rapoport DM. Detection of respiratory events during NPSG: nasal cannula/pressure. Sleep 1997;20:11751184.
Smirne S, Iannacconne S, Ferini-Strambi L, Comola M, Colombo E, Nemni
R. Muscle fibre type in habitual snoring. Lancet 1991;337:597599.
Series F, Cote C, Simoneau JA, Gelinas Y, St Pierre S, Leclerc J, Ferland R, Marc I. Physiologic, metabolic and muscle fiber type characteristics of musculus uvulae in sleep apnea/hypopnea syndrome and in
snorers. J Clin Invest 1995;95:2025.
Hudgel DW, Hendricks C. Palate and hypopharynx: sites of inspiratory
narrowing of the upper airway during sleep. Am Rev Respir Dis 1988;
138:15421547.
VOL 166
2002
61. Shepard JW Jr, Thawley SE. Localization of upper airway collapse during sleep in patients with obstructive sleep apnea. Am Rev Respir Dis
1990;141:13501355.
62. Hudgel DW, Harasick T, Katz RL, Witt WJ, Abelson TI. Uvulopalatopharyngoplasty in obstructive sleep apnea: value of preoperative localization of site of upper airways narrowing during sleep. Am
Rev Respir Dis 1991;143:942946.
63. Launois SH, Feroah TR, Campbell WN, Issa FG, Morrison D, Whitelaw
WA, Isono S, Remmers JE. Site of pharyngeal narrowing predicts
outcome of surgery for obstructive sleep apnea. Am Rev Respir Dis
1993;149:182189.
64. Schwab RJ, Gupta KB, Gefter WB, Hoffman EA, Pack AI. Upper airway soft tissue anatomy in normals and patients with sleep disordered
breathing: significance of the lateral pharyngeal walls. Am J Respir
Crit Care Med 1995;152:16731689.
9. Tests of Respiratory Muscle Function in Children

PHYSIOLOGY OF THE DEVELOPING
RESPIRATORY PUMP
In early life the ventilatory response to loaded breathing is limited. Ventilatory failure can occur in newborns and infants due to
immaturity of the chest wall and respiratory muscles, poor coupling between thoracic and abdominal movements, and upper
airway dysfunction. Furthermore, infants spend a large proportion of time asleep, with full-term newborns spending more than
50% of their time in rapid eye movement (REM) sleep and premature infants an even higher proportion of time in REM (1).
At birth, the ribs extend almost at right angles from the vertebral column. As a result, the rib cage is more circular than in
adults (2) and, consequently, lacks mechanical efficiency. In
adults, the volume of the rib cage can be increased by elevating the ribs. In infants, the ribs are already elevated, and this
may be one reason why motion of the rib cage during room air
breathing contributes little to tidal volume (VT) (3). The diaphragm appears flattened with a very wide angle of insertion
on the rib cage, resulting in the absence of an area of apposition (4). The orientation of the ribs does not change substantially until the infant assumes the upright posture. Concurrently, there is progressive mineralization of the ribs. Between
1 and 2 years of age (3), rib cage contribution to tidal breathing reaches a value that approximates that reported in adolescents during non-REM (NREM) sleep (5). Changes in shape
and structure with advancing postnatal age play a central role
in stiffening the rib cage. A high chest wall compliance (Ccw)
relative to lung compliance (CL) is an inherent characteristic of
the newborn mammal (6). In infants, outward recoil of the
chest wall is very low. Consequently, the static passive balance
of forces between the lung and the chest wall would dictate a
very small functional residual capacity (FRC). There are compelling reasons to believe that dynamic end-expiratory lung
volume in newborns and infants is substantially above the passively determined FRC. It has been shown that in newborns, in
contrast to adults, expiration is terminated at substantial flow
rates (7). In addition to foreshortened expiratory time, infants
use postinspiratory activity of the diaphragm (8) and expiratory glottic narrowing to actively slow expiration (9). Dynamic
elevation of end-expiratory lung volume above passive FRC
persists until around the end of the first year of life (10).
With growth, there is a progressive increase in the bulk of
respiratory muscles. There are also important changes in the
fiber composition, fiber size, and oxidative capacity and contraction characteristics of the diaphragm (11). Mean cross-sectional area of all fiber types increases postnatally. Maximal
pressures exerted by infants and even children are surprisingly
high compared with adults (1214). This is probably related to
the small radius of curvature of the rib cage, diaphragm, and
abdomen that, according to the Laplace relationship, converts
small tensions into relatively high pressures (12). However,
the inspiratory force reserve of respiratory muscles is reduced
in infants with respect to adults because inspiratory pressure
demand at rest is greater. High pressure demand in infants is
due to high minute ventilation and to high weight-corrected
metabolic rate (1517).
Fatigability of neonatal respiratory muscles as compared
with adult muscles remains a controversial issue. The paucity
of fatigue-resistant Type I fibers, the high proportion of faAm J Respir Crit Care Med Vol 165. pp 601609, 2002
tigue-susceptible
Type IIc fibers, and low oxidative capacity of
DOI:
10.1164/rccm.2102104
Internet
address:diaphragm
www.atsjournals.org
the neonatal
suggest that the muscle may be rela-
tively prone to fatigue. An in vivo study in rabbits found that

diaphragmatic fatigue occurred more quickly in neonatal than
adult animals (18). However, other in vitro and in vivo animal
studies have shown the opposite (19).
Chest wall muscle contraction helps to stabilize the compliant infant rib cage, minimizing inward displacement of the rib
cage by diaphragmatic contraction. However, when the stabilizing effect of intercostal muscles is inhibited, such as during
REM sleep, paradoxic inward motion of the rib cage occurs
during inspiration (20, 21). During REM sleep, the diaphragm
dissipates a large fraction of its force in distorting the rib cage
rather than effecting volume change. This increase in diaphragmatic work of breathing (22) represents a significant expenditure of calories, and may contribute to the development
of diaphragmatic fatigue and ventilatory failure. Furthermore,
acidosis and hypoxia, both of which increase muscle fatigability, are not uncommon in sick premature infants.
As in adults, upper airway muscles actively dilate and
stiffen the airway during inspiration in infants and children.
As cited above, laryngeal adduction is prominent during expiration in infants, effectively increasing the time constant for
lung emptying. Upper airway muscles, except for the nasal and
laryngeal abductors (alae nasi and posterior cricoarytenoid
muscles), become atonic during REM sleep, predisposing the
upper airway to collapse during inspiration and possibly resulting in decreased end-expiratory lung volume due to loss of
laryngeal airflow braking. The sparing of active nasal and laryngeal dilation during inspiration in REM sleep presumably
decreases the work of breathing by maintaining low airflow
resistance in the upper airway. The negative pressure required
to collapse the upper airway appears to be low in the neonate
(23) (approximately 3 cm H2O), increasing in children to
about 20 cm H2O (24) and returning to levels similar to
those in early infancy with adulthood (25). Children and
adults with obstructive sleep apnea exhibit airway collapse at
pressures at or near 0 cm H2O, suggesting that they have a
more compliant upper airway than do normal subjects (24, 25).
TESTS OF RESPIRATORY MUSCLE FUNCTION

Equipment and Measurement Conditions
Maximal respiratory muscle pressures in infants and children

are similar to those encountered in adults (see 2628). The equipment used to measure respiratory muscle and chest wall function
must be modified, however, to accommodate the smaller flow
rate and VT of infancy and childhood. The following equipment should be available: pressure transducers (flow, 2 cm
H2O; esophageal pressure [Pes] and airway opening pressure
[Pao], 50 cm H2O); transducer amplifiers; heated pneumotachographs (012 L/minute for neonates to 6 months, 030 L/
minute for 3 months to 2 years); esophageal balloons; catheters (8F, internal diameter of 2 mm or 6F, internal diameter of
1.6 mm); face masks; timed shutter with pressure port; and
multichannel strip chart recorder or computer screen, central
processing unit, analog-to-digital converter, and printer. Because Pes may not adequately represent pleural pressure in
the presence of high Ccw, all esophageal balloon and catheter
measurements in infants should be validated by an occlusion
test (29, 30). Equipment dead space should not exceed 1.5
ml/kg body weight.
For respiratory inductive plethysmography in children, the
following are required: rib cage (RC) and abdominal (AB)
602
bands of a size appropriate for the subjects chest wall; a source

of oscillatory current; signal processing equipment necessary
to produce voltage output proportional to changes in RC and
AB cross-sectional area. Also required is appropriate recording
equipment such as strip chart, oscilloscope, or analog-to-digital
converter with computer screen, central processing unit, and
printer; or Respicomp, Respitrace PT (Non-Invasive Monitoring Systems [Miami Beach, FL], with computer hardware and
analysis software integrated into system). Calibration of the
respiratory inductive plethysmograph equipment is needed to
measure VT (spirometer or pneumotachograph). Frequency
amplitude response of respiratory inductive plethysmography
has been shown to be flat to 13 Hz (31). The equipment necessary for the measurement of surface electromyograms (EMGs)
in infants and children is identical to that used in adults (see
EMG EQUIPMENT in Section 3 of this Statement).
At less than 57 years of age, children are not able to co-operate with lung function testing. For lung function testing of infants, a sedative such as chloral hydrate is usually administered.
Therefore, the laboratory should be quiet, dimly lit, and conducive to sleeping. Sedation should not be given to infants with
known upper airway obstruction. Continuous pulse oximetry
should be measured in all sedated infants. Supplemental oxygen
administration may be required, resuscitation equipment should
be available, and personnel should be trained in pediatric cardiopulmonary resuscitation. Infants should not be discharged
from the pulmonary function laboratory until fully awake.
For most tests in infants, the supine position is the position
for which normal standards are defined. The neck should be in
a neutral or a slightly extended position. The infant should be
in clinically determined quiet sleep.
For tests in older children requiring maximal efforts, such
as maximal respiratory muscle strength assessment, the technicians should be experienced in working with young children
and in striking the right balance between exhorting the child
to give his/her best effort, and not frightening the child. Other
measurement conditions are identical to those for adult pulmonary function laboratories.
Chest Wall Compliance
Rationale. The high Ccw of infants influences FRC and ability

to withstand respiratory loading. Certain disease states, e.g.,
neuromuscular disease, bony thoracic abnormalities, and intra-abdominal processes such as ascites, can alter Ccw.
Methods. Children greater than 7 years of age can probably understand the directions necessary for Ccw measurement
as described for adults in Section 6 of this Statement, although
passage of esophageal balloons in this age range is not routine.
The measurement of passive Ccw in younger children and infants is technically difficult because of their inability to cooperate. To measure passive Ccw, the respiratory pump muscles
must be inactive, or as close to inactive as possible. Several
methods to achieve this have been described:
1. Passive Ccw has been measured under quasistatic conditions, during mechanical ventilation (MV) in preterm and
full-term infants (32). Because during MV, pressure inside
the chest wall is pleural pressure (approximated by an
esophageal balloon or catheter), and pressure outside the
chest wall is atmospheric (referenced to zero), the pressure
difference across the chest wall is Pes. With an esophageal
balloon in place, measurements of Pes at points of zero flow
(end inspiration [PI], following a 1.0-second end-inspiratory
hold and end expiration [PE]) are made, and the difference
is divided into that breaths VT. Chest wall compliance is
then measured as: Ccw VT/(PI,es PE,es).
VOL 166
2002
2. Passive dynamic Ccw has been measured in spontaneously

breathing sedated infants and toddlers (33). An anesthesia
bag and mask is placed over the infants mouth and nose,
and gentle insufflations are given until the infant is no
longer making respiratory efforts, as indicated by surface
diaphragmatic EMG electrodes or by a conversion of Pes
deflections from negative to positive during inspiration.
With an esophageal catheter in place, CL is calculated as:
CL VT/[(PI,ao PI,es) (PE,ao PE,es)], where PI,ao
and PE,ao are pressures at the airway opening during inspiration and expiration, respectively.
3. Respiratory system compliance (Crs) is calculated as: Crs
VT/(PI,ao PE,ao), and Ccw is calculated as: 1/Ccw 1/
Crs 1/CL.
4. Passive static Ccw has been measured in preterm infants
(32). After an end-inspiratory occlusion, Pes rises to a plateau in parallel with Pao, because the lung is at isovolume
and transpulmonary pressure (Ptp) remains constant. After
release of the occlusion, Pes decreases as lung volume decreases. Ccw is calculated as: Ccw dV/(PI,es PE,es),
where dV is the change in lung volume.
This technique can be performed in spontaneously breathing
infants, and relies on an intact HeringBreuer reflex causing
relaxation of the chest wall after an end-inspiratory occlusion.
Advantages and disadvantages of Ccw measurements. Measurements of chest wall mechanics are useful in assessing overall respiratory function in infants, because the chest wall is
highly compliant in this age range. This leads to inefficient
chest wall motion and predisposes to fatigue. These tests should
probably be performed only in the research setting, because of
their technical difficulty and requirement for absolute chest
wall muscle relaxation to achieve interpretable results.
Normal values and alterations in disease. Chest wall compliance has been measured in preterm and full-term infants,
toddlers up to the age of 3 years, school age children and adolescents, and in young adults (see Table 1). Because different
techniques have been used in each age range, it is difficult to
compare estimates of Ccw with growth. Size correction is also
important; as with CL, Ccw must be corrected for resting lung
volume, or, if that is not available, for body weight. Another
way that Ccw can be compared between different age groups
is by comparison with CL. The chest wall stiffens progressively
with age. In preterm and full-term infants, Ccw is about three
to six times greater than CL (3236); in school age children,
Ccw is approximately twice CL (37); in adolescents and adults,
Ccw is approximately equal to CL and in the elderly, Ccw is
approximately half CL (38) (Table 1). Chest wall compliance is
abnormally increased in infants and toddlers with neuromuscular disease (39). In contrast, Ccw is abnormally decreased in
adults with neuromuscular disease, possibly secondary to joint
contractures due to years of low-VT breathing (40). Although
Ccw is presumably diminished in restrictive chest wall disorders such as scoliosis and asphyxiating thoracic dystrophy,
physiologic measurements are few. Likewise, the effects of abdominal disorders such as obesity, ascites, and prune belly
syndrome on Ccw are poorly understood.
Clinical application. Measurement of Ccw is not performed
in routine clinical practice, but has been reported in clinical research.
Maximum Inspiratory and Expiratory Pressures
Rationale. Mouth PI,max and PE,max measurements are used

to study respiratory muscle strength.
Methods. Accurate mouth PI,max and PE,max measurements can be obtained at approximately 67 years of age.
603
Children must be well instructed before the test. Pressures are

measured with the child seated and wearing a noseclip. A cylindrical mouthpiece is recommended, as PE,max has been
shown to be significantly lower when a scuba-type mouthpiece
is used instead of a cylindrical one (41). A small leak, created
by placement of a needle (1 mm in diameter by 15 cm long) in
the mouthpiece, is recommended to eliminate glottic closure
and artificially high PI,max (13, 42). Each effort should be
maintained for at least 1 second. Simultaneous measurement
of the lung volume at which maximal pressure is generated is
recommended, using either a body plethysmograph or a spirometer. PI,max measurements at residual volume (RV) and PE,max
measurements at total lung capacity (TLC) require two successive cooperation-dependent maneuvers. However, PI,max
or PE,max at the end of a normal expiration, i.e., at FRC, requires only one cooperation-dependent maneuver, thereby limiting the potential for fatigue.
Advantages. Maximum inspiratory and expiratory measurements in children are simple and noninvasive.
Disadvantages. The disadvantages of PI,max and PE,max
measurements in children are their variability and a learning
effect. When fewer than five maneuvers were performed, the
coefficient of variation (CV) was found to be 9% and was independent of maneuver, age, and sex (13). Studies in healthy
children and in children with various respiratory disorders
have shown a learning effect (43, 44). With multiple PI,max
and PE,max determinations, i.e., more than 20 attempts, pressures were significantly higher than with a short PI,max and
PE,max method, i.e., fewer than five attempts (43, 44). In routine practice, it is recommended to perform five measurements or more until two reproducible maximal values are obtained. In healthy children tested seated, PI,max and PE,max
have been shown to be independent of the thoracoabdominal
configuration assumed during the maneuver (45).
Normal values. Table 2 shows normal values of PI,max and
PE,max in children and adolescents (13, 41, 42, 46, 47). In a
number of studies the level of lung volume has not been controlled during the test. Only one study provides normal values
for PI,max at FRC. Maximum inspiratory and expiratory pressures increase with age in children (15, 43), and are lower in
girls than in boys even before puberty. However, the increase
in PI,max and PE,max with age underestimates the increase in
TABLE 1. NORMAL VALUES OF CHEST WALL COMPLIANCE

WITH AGE
First
Author
Gerhardt
Papastamelos
Reference
No.
32
33
ml/cm H2O
ml/cm H2O/kg
Preterm
Infancy
5.7 1.4
17.4 6.7
(310)
(25)
2.8 0.87
(13)
2.0 0.51
13
Sharp
Mittman
37
38
Ccw
Age Range
(yr)
516
5
8
12
16
Adulthood
2029
3039
4049
5059
6069
7079
20.1 7.7
78
106
156
184
4
4
4
3
350
250
250
250
136
210
5
3.5
3.5
3.5
2
3
Ccw/CL
(36)
(12)
(12)
net muscle force, i.e., the product of pressure and surface area
over which the pressure is applied (13).
Clinical application. Measurements of PI,max and PE,max
are useful in children and/or adolescents with neuromuscular
disorders (4850), with cystic fibrosis (5155), and with chronic
obstructive pulmonary disease (16). In the case of hyperinflation, PI,max and PE,max values have to be corrected for the
absolute volume at which measurements were made (16, 53).
PI,max and PE,max measurements may be useful in malnourished children (53). Maximum inspiratory pressure at FRC
provides an assessment of the inspiratory maximal reserve of
the respiratory muscles during quiet breathing in children (16,
55, 56). It has been proposed that inspiratory muscle strength
be assessed by the measurement of the maximal pressure obtained while breathing 5% CO2 at the time of weaning from
MV in infants and children (57).
Nasal Sniff Pressures
Nasal sniff pressure (Pnas,sn) is an established test of inspiratory muscle strength in adults (see Section 2 of this Statement).
The technique has been applied to children, normal values
have been reported (58), and this simple noninvasive measure
has considerable promise in pediatric practice.
Crying Pressures: Crying PI,max and PE,max
Rationale. Mouth pressures generated during crying efforts

may provide an index of respiratory muscle strength in awake
infants.
Measurements. Firm application of a rubber cushion mask
against the face is sufficient to elicit crying efforts in the awake
infant (14). An artificial leak in the mask has been used to prevent glottic closure. Airway occlusions are performed at the end
or at the beginning of a crying effort to measure crying PI,max
and PE,max, respectively. Only peak crying PI,max without a
pressure plateau is available during crying. Crying PE,max can
be measured at a pressure plateau sustained for 1 second.
Advantages. The main advantage is simplicity.
Disadvantages. Disadvantages include the following: (1)
no determination of lung volume during the test; (2) gross potential pressure leaks around the mask; (3) artificially high
peak crying PI,max if no needle leak in the mask to prevent
glottic closure; and (4) relatively high CVs, 11 4 and 9 3%
(mean SD), respectively, as shown in one group of healthy
infants.
Normal values. Normal values have been provided in a
large group of healthy infants from the first month to the end
of the second year of life (14). Peak crying PI,max was 118
21 cm H2O and independent of age and sex. However, overestimation of peak crying PI,max may have occurred because
glottic closure was not prevented. Crying PE,max was 125 35
cm H2O and was related to body weight.
Clinical application. Crying pressures may be useful in infants with neuromuscular disease (59).
Crying Pressures: Crying Transdiaphragmatic Pressure
2
1.2
1
0.8
0.7
0.7
Definition of abbreviations: Ccw chest wall compliance; CL lung compliance.

Values are means or means SD. Values in parentheses are ranges.
Rationale. Crying transdiaphragmatic pressure (Pdi) measurements allow assessment of diaphragm muscle strength during
inspiratory crying efforts in infants.
Measurements. Accurate recordings of Pes and gastric pressure (Pga) are essential.
Advantages. Transdiaphragmatic pressure is specific for diaphragmatic contraction.
Disadvantages. Disadvantages include (1) the fact that reliability of Pdi requires accurate measurements of Pes; and (2)
similar limitations to crying PI,max with respect to variability
in lung volume at which Pdi is measured during crying efforts.
604
TABLE 2. NORMAL VALUES OF MAXIMAL INSPIRATORY AND

EXPIRATORY PRESSURES IN HEALTHY CHILDREN
AND ADOLESCENTS
Reference
Age
(yr)
PI,max (cm H2O)

At RV
At FRC
PE,max (cm H2O)

At TLC
At FRC
Males
47
13
47
13
47
13
41*
46
47
42
41
8
77 24
70 24
105 27
97 22
114 27
130 16
111 34
105 23
10
1113
1317
107 26
126 22
79
99 23
90
123 27
90101
161 37
176 24
131 30
107129
114 35
166 44
83 16
100 23
132 33
8
71 29
59 21
71 29
59 21
108 29
112 20
85 28
98 25
10
1113
1317
76 25
109 21
62
74 25
90
74 25
7786
126 32
138 31
95 29
86107
86 22
135 29
2002
tion of sleep state. There is a relatively high CV, 11.7% in children (15); variability has not been determined in infants.
Normal values. Occlusion pressure was found to be 4.4 cm
H2O in full-term newborns (61, 62) and 3.6 cm H2O in preterm
infants (62). Occlusion pressures are available in awake children from 416 years of age during resting breathing of room
air (Table 3) (15). Pmo,0.1 decreases as a power function with
age and reaches adult values at approximately 13 years of age.
Clinical application. Measurements of Pmo,0.1 are useful
to assess respiratory drive in infants and children with chronic
intrinsic loaded breathing (bronchopulmonary dysplasia [62],
interstitial lung disease [63], chronic obstructive lung disease
[16, 51, 64]).
Inspiratory Pressure Reserve and TensionTime Index
Females
47
13
47
13
47
13
41*
46
47
42
41
VOL 166
65 18
65 18
101 26
Definition of abbreviations: FRC functional residual capacity; PE,max maximum

expiratory pressure; PI,max maximum inspiratory pressure; RV residual volume;
TLC total lung capacity.
* Cylindric mouthpiece.
Mean value SD.
Mean values for 11, 13, and 17 years.
Normal values. Transdiaphragmatic pressure during crying

efforts has been measured in 18 healthy infants from 8 to 21
months postconceptional age, i.e., gestational age plus postnatal age (60). Crying Pdi is 60 cm H2O at 1 month postnatal
age, which is much lower than reported crying PI,max (14).
Crying Pdi increases with postconceptional age (60).
Clinical application. Crying Pdi may be useful in infants
with abdominal wall defects or after repair of diaphragmatic
hernia (60).
Occlusion Pressure
Rationale. Occlusion pressure (Pmo,0.1s) allows estimation of

the respiratory drive (see Section 1 of this Statement).
Measurements. Measurements of Pmo,0.1 can be performed in sleeping supine infants (61, 62) and in awake seated
children as soon as the child is able to breathe quietly with a
noseclip and a mouthpiece, i.e., after 34 years of age (15).
Even light sedation is not recommended in young children because it will affect the level of consciousness. A period of at
least 5 minutes of regular breathing is required before performing occlusions at intervals of at least 1 minute. A mean of
five acceptable occlusions is used to determine Pmo,0.1.
Advantages. Occlusion pressure is easily measured.
Disadvantages. It may be argued that the duration of pressure measurement over 100 milliseconds, as established in
adults, is not appropriate in infants and children because of a
shorter inspiration time (TI) than in adults. However, Pmo,0.1
has been shown to be significantly related to the mean inspiratory flow (VT/TI) in children from 4 to 16 years of age (15).
States of alertness influence respiratory drive and therefore
accurate measurements in infants should include determina-
Rationale. Measurement of inspiratory pressure reserve, i.e.,

the ratio of the mean inspiratory pressure (PI) during resting
breathing over PI,max at FRC, assesses the potential limitation of the inspiratory muscles to generate pressure at the endexpiratory level.
Measurements. Inspiratory pressure reserve can be assessed
in children by noninvasive mouth pressure measurements. Inspiratory pressure is calculated as 0.5 a TI (where a is
Pmo,0.1 10) (16). Inspiratory power for breathing at rest
can be calculated as PI VT/TI TI/total inspiratory time (Ttot)
(16). The critical inspiratory power (65), i.e., the inspiratory
power above which fatigue occurs, can be calculated as PI,max
VT/TI TI/Ttot. Critical PI,max is assumed to be equal to
60% of PI,max generated at FRC (65). A noninvasive pressuretime index for all the inspiratory muscles (PTImus) can
be estimated as PTImus PI/PI,max TI/Ttot (16, 55, 56).
Advantages. Assessment of inspiratory muscle function requires only noninvasive measurements. PTImus assesses all
the inspiratory muscles.
Disadvantages. The validity of the estimation of PI cannot
be proved over the entire TI. The rise in pressure during inspiration is approximated by a single power function of time assuming a linear applied pressure profile. If the inspiratory
driving pressure increases as an exponential function of time,
PI is overestimated by extrapolating the Pmo,0.1 over the entire TI. Therefore, all derived parameters including PI may be
affected by an overestimation. However, in healthy adults and
in adults with chronic obstructive pulmonary disease, PI has
been shown to be significantly correlated with Pdi, and PTImus
with the tensiontime index of the diaphragm (TTIdi) (66).
No comparison of PTImus and PTIdi is available in children.
Normal values. Table 3 shows that PI/PI,max at FRC decreases with age in children. TTImus decreases because of a decrease in PI at rest and an increase in PI,max at FRC, with no
change in TI/Ttot (15). However, no study has defined the critical PTImus above which fatigue occurs in children, as has been
done in adults for PTIdi using electromyographic criteria (67).
Clinical application. Such measurements are helpful in children with increased load of breathing to estimate the degree
of reduction of their inspiratory pressure reserve, the inspiratory power for breathing at rest, and the breathing strategy required to remain below the fatigue threshold of the respiratory muscles. It has been shown that the greater the PI/PI,max
at FRC, the lower the TI/Ttot in children with chronic obstructive pulmonary disease (16). The reduction in inspiratory pressure reserve limits the possibilities of adaptation of the pattern
of breathing when an increase in the breathing load occurs, such
as during exercise (68) or during sleep. The younger the child,
the greater the risk (64). Such measurements should be performed in the follow-up of severe respiratory disorders, such as
during the wait for lung transplant.
605
25.6
14.7
18.9
13.0
0.07
0.09
0.06
costal muscles are inhibited during REM sleep and chest wall
distortion occurs during inspiration. No abdominal muscle activation has been reported in healthy infants nursed in the supine position during NREM sleep (71). Activation of abdominal muscles is observed during NREM sleep while breathing
CO2, but not during REM sleep (71).
Clinical application. The main utility of recording of EMG
of the respiratory muscles is during sleep studies. Persistent
phasic chest wall EMG activity helps to identify obstructive
respiratory events, especially obstructive hypopneas during
sleep, and therefore should limit the use of the invasive measurement of Pes (76). EMG recording of abdominal muscles
assesses the degree of abdominal muscle contraction in children with loaded breathing related to partial upper airway obstruction during sleep (71, 72).
13.9
0.06
Integrated Pump Function: Thoracoabdominal Motion
TABLE 3. NORMAL VALUES OF OCCLUSION PRESSURE,

INSPIRATORY FORCE RESERVE, AND PRESSURETIME INDEX
IN HEALTHY CHILDREN AND ADOLESCENTS
Age
(yr)
4
6
8
Sex
Pmo,0.1*
(cm H2O)
PI
(cm H2O)
PI,max,FRC
(cm H2O)
Male
Female
Male
Female
Male
3.60
3.60
2.80
2.80
2.34
15.2
70
Female
2.34
15.2
59
Male
Female
Male
Female
2.04
2.04
1.82
1.82
14.3
14.3
13.7
13.7
97
77
105
98
Male
Female
Male
Female
1.65
1.65
1.52
1.52
PI/PI,max,FRC
(%)
21.7
10
12
14
16
PTImus
0.10
0.11
Definition of abbreviations: Pmo,0.1 occlusion pressure; PI mean inspiratory pressure; PI,max,FRC maximal inspiratory pressure generated at functional residual capacity; PTImus pressuretime index for all the inspiratory muscles.
* Pmo,0.1 (occlusion pressure) from Gaultier and coworkers (15); Pmo,0.1 is not significantly different between males and females.
PI,max,FRC (maximal inspiratory pressure generated at functional residual capacity)

from Gaultier and Zinman (13).
TI/Ttot does no change with age in children (15), TTImus was calculated using a T I/
Ttot equal to 0.45, i.e., the mean value obtained in children between 4 and 16 years of
age.
Electromyography of the Respiratory Muscles
Rationale. Electromyography can be used to assess the level

and pattern of activation of the respiratory muscles.
Measurements. Surface electrodes have been used nearly
exclusively for EMG recording of the respiratory muscles in
infants and children. Surface electrodes are placed in the right
sixth and seventh interspaces between the midaxillary and midclavicular lines for recording of diaphragmatic EMG (EMGdi)
(6971), in the second intercostal space parasternally for intercostal EMG recording, and at the midpoint between the umbilicus and iliac crest for abdominal muscle recording (71, 72).
One study reported EMGdi using an esophageal electrode
with specific equipment adapted for use in sleeping infants
(73). No studies have been published using intramuscular electrodes for recording EMG of the respiratory muscles in infants or children.
Advantages. The recording of EMG of the respiratory
muscles using surface electrodes is noninvasive.
Disadvantages. Surface electrode EMGdi tracings can be
contaminated by other inspiratory or expiratory muscles (73).
Therefore, when recording diaphragmatic EMG activity with
surface electrodes, it is more appropriate to describe the values as chest wall EMG rather than EMGdi. Techniques to remove cardiac artifacts on the EMGdi have to take into account
the high cardiac frequency in infants, especially in preterm infants (74) (see ARTIFACTS in Section 3 of this Statement).
Normal values. Quantitative measurements of chest wall
EMG recorded with surface electrodes have been reported for
EMG time and frequency domains (69, 70, 74, 75). Few studies have described the activation of the respiratory muscles
during quiet breathing in infants without respiratory disorders. In infants nursed in the supine position, phasic and tonic
activity of the intercostal muscles have been reported during
NREM sleep (69). Contraction of the intercostal muscles stabilizes the highly compliant chest wall of the infant. The inter-
Rationale. The respiratory muscles (diaphragm and intercostals) act on the rib cage to effect respiratory motion and ventilation. Rib cage motion can be taken as an index of intercostal
muscle action, while AB motion can be taken as an index of
diaphragmatic descent. Thus thoracoabdominal motion (TAM)
provides a visual index of respiratory muscle function.
Methods. The most widely used method to assess TAM is respiratory inductive plethysmography, although strain gauges and
magnetometers are also used. Respiratory inductive plethysmography uses thin cloth bands that are placed around the RC
and AB. A wire is sewn into the bands, and when an oscillatory
current is applied, changes in the cross-sectional area of the
chest wall are reflected as changes in the electric inductance of
the wires, displayed as a change in voltage output. Respiratory
inductive plethysmography can be used to quantitate asynchrony by measuring a phase angle, ( 0, synchronous
breathing; 180, paradoxic breathing), between the RC and
AB compartments. The calculation of does not require calibration for volume. Respiratory inductive plethysmography can
also be used to quantitate the relative contribution of the RC
and AB to VT; this does require calibration for volume. Several
techniques have been described for volumetric calibration. The
isovolume, least mean squares, and quantitative diagnostic calibration techniques are described elsewhere (7781) (see DEVICES
USED TO MONITOR BREATHING in Section 6 of this Statement).
Advantages and disadvantages of respiratory inductive plethysmography. Respiratory inductive plethysmography is valuable as a direct, noninvasive measure of chest wall motion and
an indirect measure of respiratory muscle function. Being an
indirect measure, however, chest wall motion can reflect
events other than respiratory muscle function. Chest wall motion is a final common pathway of integrated respiratory system output; it can be influenced by underlying lung mechanics
and Ccw. Thus, asynchronous chest wall motion can represent
neuromuscular weakness (8284), fatigue (85), high Ccw (86),
abnormally low CL or high lung resistance (87, 88), upper airway obstruction (89), effects of anesthesia (90), or a combination of two or more of these. Its major weakness is thus that it
is a nonspecific indicator, and abnormal chest wall motion
must be interpreted in the context in which it occurs.
Calibration of respiratory inductive plethysmography for
volume is difficult in infants, especially in preterm infants with
highly compliant chest walls and paradoxic chest wall motion.
The highly compliant infant chest wall probably invalidates
the assumption of two degrees of freedom (RC and AB) that
is the basis for most respiratory inductive plethysmography
calibration equations. Furthermore, phase angles may be difficult to interpret if breathing does not approximate a sinusoidal pattern (91).
606
Normal values and alterations in disease. In infants, the high

Ccw predisposes to asynchronous chest wall motion. Thus, normal preterm infants display asynchronous chest wall motion
during NREM sleep (86). Normal full-term infants display synchronous RCAB motion in NREM sleep ( 8 7 [mean
SD]) (84), but may display asynchronous motion in REM sleep
(21). The time spent with paradoxic inward RC movement during REM sleep decreases with advancing age; paradoxic, RC
motion is present during nearly 100% of the REM time in the
newborn, but only 10% of the REM time by 3 years of age
(21). For these reasons, it should be noted whether the infant is
in the quiet awake state, or in clinically determined quiet or active sleep, when performing measurements of TAM.
Normal adults and children have synchronous breathing
between the RC and AB compartments (92, 93).
As far as neuromuscular disease is concerned, assessment
of TAM can pinpoint the site of weakness: paradoxic inward
motion of the RC during inspiration indicates intercostal muscle weakness, whereas paradoxic inward motion of the AB during inspiration indicates diaphragmatic weakness (82, 83). Such
chest wall asynchrony disappears during MV (94, 95), and may
indicate respiratory muscle rest.
Turning to chronic airflow obstruction, in adults, abnormalities of diaphragmatic function accompany chronic hyperinflation (96). Similar problems occur in infancy and childhood. Preterm infants develop chest wall asynchrony and
decreased minute ventilation when breathing through inspiratory resistive loads (8); it is not clear whether this represents
respiratory muscle fatigue or simply high Ccw in the face of increasing negative intrathoracic pressure. Infants with bronchopulmonary dysplasia display asynchronous RCAB motion. The degree of asynchrony is proportional to the degree
of abnormality in lung compliance and resistance (88).This improves on administration of aerosolized bronchodilators (87).
Upper airway obstruction can likewise cause thoracoabdominal asynchrony (89).
Clinical application. Measurements of TAM have been
made in clinical research studies as described. By differentiating the summed RC and AB signals, tidal flowvolume curves
have been analyzed (97). Measurements of TAM are also used
in routine clinical practice, particularly for detecting upper airway obstruction during sleep studies. They are also used in infant monitoring devices to detect apnea.
The following two methods of analyzing integrated respiratory pump function have limited clinical applications at present
and are primarily techniques under investigation.
1. Diaphragmatic movement (real-time ultrasonography): Diaphragmatic excursions have been measured in full-term
infants by real-time ultrasonography (98). Only the right
hemidiaphragm is accessible by this technique, through the
ultrasonographic window provided by the liver; the left
hemidiaphragm is obscured by stomach and bowel gas. The
posterior aspect of the diaphragm moves to a greater extent than the anterior aspect, perhaps because of the effect
of the larger posterior area of apposition to the inner chest
wall. Diaphragmatic excursions are lessened, and the diaphragm moves more uniformly, in infants who are paralyzed and undergo MV. Similar observations (4) led to the
conclusion that, unlike the piston-like motion of the adult
diaphragm, the infant diaphragm has a bellows-like motion. Normal ultrasonographic values for diaphragmatic
excursion in infants and adults have been presented (4, 99
101); there is a wide degree of variation between the infant
studies, suggesting that, at least for now, clinical application is limited (Table 4).
VOL 166
2002
2. Phrenic nerve stimulation (see STIMULATION TESTS in Section

3 of this Statement): Transcutaneous phrenic nerve stimulation has been applied to children in the clinical setting; to
date phrenic nerve latency has been studied as a way of assessing phrenic nerve damage in, for example, patients recovering from cardiac surgery, but no studies of diaphragmatic fatigue have been performed. Normal phrenic nerve
latency is on the order of 4.56.5 milliseconds. There is less
than 10% day-to-day variability of this measurement (102).
Phrenic nerve latency time diminishes slightly with age, by
0.5 to 1 millisecond between birth and 10 years (103). Delays of 2 milliseconds or longer may indicate phrenic nerve
damage (102). Magnetic cervical stimulation may be an alternative, painless way to stimulate the phrenic nerves (104).
Anterior magnetic phrenic nerve stimulation is an established technique in adults (see Section 2 of this Statement).
Experience with the technique in children is limited but the
method can be used to assess diaphragm function in neonates in
the intensive care unit (109).
Upper Airway Muscles
(See ELECTROMYOGRAPHY in Section 8 of this Statement.)

Rationale. Assessment of upper airway muscle activity and
airway collapsibility is useful for assessing infants and children
with increased upper airway resistance.
Measurements. Surface and fine wire EMGs have been
used in research settings to evaluate the activity of upper airway muscles including the intrinsic laryngeal muscles (8), alae
nasi (105), and the genioglossus (105). The function of upper
airway muscles is normally assessed clinically by visualization
of the size of the airway using fiberoptic endoscopy (106), fluoroscopy (107), or, more recently, fast magnetic resonance imaging techniques (108). Another useful indirect measurement,
especially when assessing patients for obstructive sleep apnea,
is to monitor airflow at the mouth/nose for evidence of reduced flow (109). Upper airway collapsibility can be assessed
either by monitoring pharyngeal pressure during an airway occlusion or by applying positive and negative pressures at the
airway opening and looking for evidence of airflow limitation
during spontaneous breaths (24). The critical pressure (Pcrit)
of the upper airway is defined from such measurements as the
pressure at which flow equals zero as extrapolated from a
maximum flow versus applied nasal pressure curve. Closing
pressure of the upper airway has been measured in postmortem infants (110) and in normal sleeping infants (23).
Advantages. Measurement of upper airway muscle activity
provides valuable information when assessing infants and children with suspected upper airway dysfunction. Visualization
of the upper airway is a standard clinical procedure.
Disadvantages. Electromyographic measurements are not
feasible for clinical assessment because of the technical difficulties, limited number of muscles that are accessible to surface measurements, and the inability to compare levels of EMG
activity between patients. At the present time, measures of collapsibility are time-consuming and require special expertise.
Normal values. No normal values exist for the EMG of the upper airway muscles. In general, abductor muscles are active during
inspiration with quiet breathing and some adductors are active
during expiration. Measurements of Pcrit in a small number of
children yielded values of 19.5 cm H2O for children with primary
snoring and 1.0 cm H2O in children with obstructive sleep apnea
(24). Closing pressures in postmortem infants averaged 0.8 cm
H2O (110) whereas those in sleeping infants averaged 3.8 cm
H2O (23). The more positive value in the postmortem measurements probably reflects the lack of upper airway muscle activation.
607

TABLE 4. AXIAL DIAPHRAGM DISPLACEMENT: NORMAL
ULTRASONOGRAPHIC VALUES DURING TIDAL BREATHING
11.
Axial Diaphragm Displacement (mm)
Reference
Anterior
Middle
Posterior
Mean
97
4
98
2.6
7.4
3.7
4.6
12.1
3.6
9.7
15.2
Infancy
Adulthood
12.
13.
Clinical application. Measurements of upper airway function are useful in assessing children with suspected obstructive
sleep apnea due to abnormalities of either structure or upper
airway muscle activation including those with craniofacial abnormalities, hypertrophied lymphoid tissue, and neuromuscular disease. They can also be useful in cases of vocal cord dysfunction.
14.
15.
16.
17.
CONCLUSION
18.
The majority of techniques and tests described in this Section

of the Statement are not presently in routine clinical use and
many should be considered experimental. The tests can be put
into one of three groups based on their feasibility in different
settings.
19.
1. Tests that are well established and validated for routine

clinical use include PI,max, PE,max and the direct visualization of upper airway structures.
2. Many tests are fairly well developed but still require validation and establishment of normal values. These tests are
available for use in clinical investigations where there are
specifically trained personnel and specialized equipment.
They include measurements of Pmo,0.1, inspiratory pressure reserve and PTI, thoracoabdominal asynchrony, EMG
of the respiratory and upper airway muscles, and critical
and closing pressures of the upper airway.
3. Tests that are still under development and that are recommended for use only in research settings include measurement of Ccw, crying pressures, ultrasonography of the diaphragm, and phrenic nerve stimulation. These tests all require
further investigation before their feasibility in clinical practice
can be established.
20.
21.
22.
23.
24.
25.
26.
27.
28.
References
1. Curzi-Dascalova L, Peirano P, Morel-Kahn F. Development of sleep
states in normal premature and full-term newborns. Dev Psychobiol
1988;2:431444.
2. Openshaw P, Edwards S, Helms P. Changes in rib cage geometry during childhood. Thorax 1984;39:624627.
3. Hershenson M, Colin A, Wohl M, Stark A. Change in the contribution
of the rib cage to total breathing during infancy. Am Rev Respir Dis
1990;141:922925.
4. Devlieger H, Daniel H, Marchal G, Moerman P, Casaer P, Eggermont
E. The diaphragm of the newborn infant: anatomical and ultrasonographic studies. J Dev Physiol 1991;16:321329.
5. Tabachnik E, Muller N, Bryan A, Levison H. Changes in ventilation
and chest wall mechanics during sleep in normal adolescents. J Appl
Physiol 1981;51:557564.
6. Agostoni E. Volumepressure relationships to the thorax and lung in
the newborn. J Appl Physiol 1959;14:909913.
7. Kosch P, Stark A. Dynamic maintenance of end-expiratory lung volume in full-term infants. J Appl Physiol 1984;57:11261133.
8. Kosch P, Hutchison A, Wozniak J, Carlo W, Stark A. Posterior cricoarytenoid and diaphragm activities during tidal breathing in neonates. J Appl Physiol 1988;664:19681978.
9. Harding R, Jonhson P, McClelland M. Respiratory function of the larynx in developing sheep and the influence of sheep state. Respir
Physiol 1980;40:165170.
10. Colin A, Wolh M, Mead J, Patken J, Glass G, Stark A. Transition from
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
dynamically maintained to relaxed end-expiratory volume in human

infants. J Appl Physiol 1989;67:21072111.
Sieck G, Fournier M. Developmental aspects of diaphragm muscle cell,
structural and functional organization. In: Haddad G, Farber J, editors. Developmental neurobiology of breathing. New York: Marcel
Dekker; 1991. p. 375428.
Cook C, Mead J, Ozales M. Static volumepressures characteristics of
the respiratory system during maximum effort. J Appl Physiol 1964;
19:10161022.
Gaultier C, Zinman R. Maximal static pressures in healthy children.
Shardonofsky F, Perez-chada D, Carmuega E, Milic-Emili J. Airway
pressure during crying in healthy infants. Pediatr Pulmonol 1989;6:1418.
Gaultier C, Perret L, Boule M, Buvry A, Girard F. Occlusion pressure
and breathing pattern in healthy children. Respir Physiol 1981;46:7180.
Gaultier C, Boule M, Tournier G, Girard F. Inspiratory force reserve
of the respiratory muscles in children with chronic obstructive pulmonary disease. Am Rev Respir Dis 1985;131:811815.
Milic-Emili J. Respiratory muscle fatigue and its implication in respiratory distress syndrome. In: Cosmi E, Scaepelli E, editors. Pulmonary
surfactant system. Rome, Italy: Elsevier Science; 1983. p. 135141.
Lesouef P, England S, Stogryn H, Brayn A. Comparison of diaphragmatic
fatigue in newborn and older rabbits. J Appl Physiol 1988;65:10401044.
Sieck G, Fournier M, Blanco C. Diaphragm muscle fatigue resistance
during postnatal development. J Appl Physiol 1991;71:458464.
Curzi-Dascalova L. Thoraco-abdominal respiratory correlations in infants: constancy and variability in different sleep states. Early Hum
Dev 1978;2:2538.
Gaultier C, Praud J, Canet E, Delapeche M, Dallest A. Paradoxical inward rib cage motion during rapid eye movement sleep in infants
and young children. J Dev Physiol 1987;9:391397.
Guslits B, Gaston S, Bryan M, England S, Bryan A. Diaphragmatic
work of breathing in premature human infants. J Appl Physiol
1987;62:14101415.
Roberts J, Reed W, Mathew O, Menon A, Thach B. Assessment of
pharyngeal airway stability in normal and micrognathic infants. J
Appl Physiol 1985;58:290299.
Marcus C, McColley S, Carroll J, Loughlin G, Smith P. Upper airway
collapsibility in children with obstructive sleep apnea syndrome. J
Appl Physiol 1994;77:918924.
Gleadhill I, Schwartz A, Schubert N, Wise R, Permutt S, Smith P. Upper airway collapsibility in snorers and in patients with obstructive
hypopnea and apnea. Am Rev Respir Dis 1991;143:13001303.
American Thoracic Society/European Respiratory Society. Respiratory muscle function in infants: measurement conditions. Am J Respir
Crit Care Med 1995;151:20582064.
Gaultier C, Fletcher M, Beardsmore C, England S, Motoyama E. Respiratory function measurements in infants: measurement conditions. Eur Respir J 1995;8:10571066.
Sly P, Davis G. Equipment requirements for infant respiratory function
testing. In: Stocks J, Sly P, Tepper R, Morgan W, editors. Infant respiratory function testing. New York: Wiley-Liss; 1996. p. 4580.
Beardsmore C, Helms P, Stocks J, Hatch D, Silverman M. Improved
esophageal balloon technique for use in infants. J Appl Physiol 1980;
49:735742.
Lesouef P, Lopes J, England S, Bryan H, Bryan A. Influence of chest
wall distortion on esophageal pressure. J Appl Physiol 1983;55:353.
Adams J. Respiratory inductive plethysmography. In: Stocks J, Sly P,
Tepper R, Morgan W, editors. Infant respiratory function testing.
New York: Wiley-Liss; 1996. p. 139164.
Gerhardt T, Bancalari E. Chest wall compliance in full term and premature infants. Acta Pediatr Scand 1980;69:359364.
Papastamelos C, Panitch H, England S, Allen J. Developmental
changes in chest wall compliance during infancy and early childhood.
Davis G, Coates A, Papageorgiou A, Bureau M. Direct measurement
of static chest wall compliance in animal and human neonates. J Appl
Physiol 1988;65:10931098.
Reynolds R, Etsten B. Mechanics of respiration in apneic anesthetized
infants. Anesthesiology 1966;27:1319.
Richard C, Bachman L. Lung and chest wall compliance in apneic paralyzed infants. J Clin Invest 1961;40:273278.
Sharp J, Druz W, Balagot R, Bandelin V, Danon J. Total respiratory
compliance in infants and children. J Appl Physiol 1970;29:775779.
Mittman C, Edelman N, Norris A, Shock N. Relationship between
chest wall and pulmonary compliance and age. J Appl Physiol 1965;
20:12111216.
608
39. Papastamelos C, Panitch H, Allen J. Chest wall compliance in infants

and children with neuromuscular disease. Am J Respir Crit Care Med
1996;154:10451048.
40. Estenne M, Heilporn A, Delhez L, Yelnault J, De Troyer A. Chest wall
stiffness in patients with chronic respiratory muscle weakness. Am
Rev Respir Dis 1983;128:10021007.
41. Szeinberg A, Marcotte J, Roizin H, Mindotff C, England S, Tabachnik
E, Levison H. Normal values of maximal inspiratory and expiratory
pressures with a portable apparatus in children, adolescents, and
young adults. Pediatr Pulmonol 1987;3:255258.
42. Smyth R, Chapman K, Rebunk A. Maximal inspiratory and expiratory
pressures in adolescents: normal values. Chest 1984;86:568572.
43. Wagener J, Hibbert M, Landau L. Maximal respiratory pressures in
children. Am Rev Respir Dis 1984;129:873875.
44. Wen A, Woo M, Keens T. How many attempts are required to accurately measure maximal inspiratory pressures? Am J Respir Crit
Care Med 1996;153:A559.
45. Zinman R, Gaultier C. Abdominothoracic configuration and maximal
static pressures in children. Respir Physiol 1984;55:3946.
46. Leech J, Ghezzo H, Stevens D, Becklale M. Respiratory pressures and
function in young adults. Am Rev Respir Dis 1983;128:1723.
47. Wilson S, Cooke N, Edwards R, Spiro S. Predicted normal values for
maximal respiratory pressures in Caucasian adults and children.
Thorax 1984;39:535538.
48. Begin R, Bureau M, Lupien L, Lemieux B. Control and modulation of
respiration in Steinerts myotonic dystrophy. Am Rev Respir Dis 1980;
121:281289.
49. Inkley S, Oldenburg F, Vignos P. Pulmonary function in Duchenne
muscular dystrophy related to stage of disease. Am J Med 1974;56:
297306.
50. Lands L, Zinman R. Maximal static pressure and lung volumes in a
child with Guillain-Barr syndrome ventilated by a cuirass respirator. Chest 1986;89:757760.
51. Bureau M, Lupien L, Begin R. Neural drive and ventilatory strategy of
breathing in normal children, and patients with cystic fibrosis and
asthma. Pediatrics 1981;68:187194.
52. Szeinberg A, England S, Mindorff C, Fraser I, Levison H. Maximal inspiratory and expiratory pressures are reduced in hyperinflated, malnourished young adult male patients with cystic fibrosis. Am Rev
Respir Dis 1985;32:766769.
53. Lands L, Desmond K, Demizio D, Pavilanis A, Coates A. The effects of
nutritional status and hyperinflation on respiratory muscle strength
in children and young adults. Am Rev Respir Dis 1990;141:15061509.
54. Lands L, Heigenhauser G, Jones N. Respiratory and peripheral muscle
function in cystic fibrosis. Am Rev Respir Dis 1993;147:865869.
55. Hayot M, Guillaumont S, Ramonatxom M, Voisin M, Prefaut C. Determinants of the tensiontime index of respiratory muscles in children
with cystic fibrosis. Pediatr Pulmonol 1997;23:336343.
56. Gaultier C. Tensiontime index of respiratory muscles in children [editorial]. Pediatr Pulmonol 1997;23:327329.
57. Gozal D, Shoseyov D, Keens T. Inspiratory pressures with CO2 stimulation and weaning from mechanical ventilation in children. Am Rev
Respir Dis 1993;147:256261.
58. Stefanutti D, Fitting JW. Sniff nasal inspiratory pressure: reference values in Caucasian children. Am J Respir Crit Care Med 1999;159:107111.
59. Shardonofsky F, Perez-chada D, Milic-Emili J. Airway pressures during crying: an index of respiratory muscle strength in infants with
neuromuscular disease. Pediatr Pulmonol 1991;10:172177.
60. Scott C, Nickerson B, Sargen TC, Platzker A, Warburton D, Keens T.
Developmental pattern of maximal transdiaphagmatic pressure in
infants during crying. Pediatr Res 1983;17:707709.
61. Fisher J, Mortola J, Smith B, Fox G, Weeks S. Respiration in newborns: development of the control of breathing. Am Rev Respir Dis
1982;125:650657.
62. Greenspan J, Wolfson M, Locke R, Allen J, Shaffer T. Increased respiratory drive and limited adaptation to loaded breathing in bronchopulmonary dysplasia. Pediatr Res 1992;32:356359.
63. Gaultier C, Perret L, Boule M, Tournier G, Girard F. Control of
breathing in children with interstitial lung diseases. Pediatr Res 1982;16:
779783.
64. Gaultier C, Perret L, Boule M, Baculard A, Grimfeld A, Girard F. Occlusion pressure and breathing pattern in children with chronic obstructive
pulmonary disease. Bull Eur Physiopathol Respir 1982;18:851862.
65. Roussos C, Fixley M, Gross D, Macklem P. Fatigue of inspiratory muscles and their synergic behavior. J Appl Physiol 1979;46:897904.
66. Ramonatxo M, Boulard P, Prefaut C. Validation of a noninvasive tensiontime index of inspiratory muscles. J Appl Physiol 1995;78:646653.
VOL 166
2002
67. Bellemare F, Grassino A. Evaluation of the human diaphragmatic fatigue. J Appl Physiol 1982;53:11961206.
68. Boule M, Gaultier C, Girard F. Breathing pattern during exercise in
untrained children. Respir Physiol 1989;75:225234.
69. Muller N, Volgyesy G, Becker L, Bryan A. The consequence of diaphragmatic muscle fatigue in the newborn infant. J Pediatr 1979;95:793797.
70. Muller N, Volgyesi G, Becker L, Bryan M, Bryan A. Diaphragmatic
muscle tone. J Appl Physiol 1979;47:279284.
71. Praud J, DAllest A, Nedelcoux H, Curzi-Dascalova L, Guilleminault
C, Gaultier C. Sleep-related abdominal behavior during partial or
complete obstructed breathing in prepubertal children. Pediatr Res
1989;25:347350.
72. Jeffries B, Brouillette R, Hunt C. Electromyographic study of some accessory muscles of respiration in children with obstructive sleep apnea. Am Rev Respir Dis 1984;129:696702.
73. Reis F, Cates D, Landriault L, Rigatto H. Diaphragmatic activity and
ventilation in preterm infants. Biol Neonate 1994;65:1624.
74. Chambille B, Vardon G, Monrigal J, Dehan M, Gaultier C. Technique
of on-line analysis of diaphragmatic electromyogram activity in the
newborn. Eur Respir J 1989;2:883886.
75. Nugent S, Finley J. Spectral analysis of the EMG and diaphragmatic
muscle fatigue during periodic breathing in infants. J Appl Physiol
1985;58:830833.
76. American Thoracic Society. Standards and indications for cardiopulmonary sleep studies in children. Am J Respir Crit Care Med 1996;
153:866878.
77. Konno K, Mead J. Measurement of the separate volume changes of the rib
cage and abdomen during breathing. J Appl Physiol 1967;22:407422.
78. Dolfin T, Duffy P, Wilkes D, Bryan M. Calibration of respiratory inductive plethysmography. Am Rev Respir Dis 1982;125:577579.
79. Stefano J, Spizer A, Baumgart S, Davis J, Fox W. Inductive plethysmography: a facilitated postural calibration technique for rapid and
accurate tidal volume determination in low birth weight premature
newborns. Am Rev Respir Dis 1986;134:10201024.
80. Sackner M, Warson H, Belsito A, Feinerman D, Suarez M, Gonzalez
G, Bizousky F, Krieger B. Calibration of respiratory inductive plethysmograph during natural breathing. J Appl Physiol 1989;66:410420.
81. Revow M, England S, Stogryn H, Wilkes D. Comparison of calibration
methods for respiratory inductive plethysmography on infants. J
Appl Physiol 1987;63:18531861.
82. Mortola J, SantAmbrogio G. Motion of the rib cage and the abdomen
in tetraplegic patients. Clin Sci Mol Med 1978;54:2532.
83. DeTroyer A, Kelly S. Chest wall mechanics in dogs with acute diaphragm paralysis. J Appl Physiol 1982;53:373379.
84. Higgenbottam T, Allen D, Loh LC, Clark TJ. Abdominal wall movement in normals and patients with hemidiaphragmatic and bilateral
diaphragmatic palsy. Thorax 1977;32:589595.
85. Roussos C, Macklem P. The respiratory muscles. N Engl J Med 1982;
307:786797.
86. Deoras K, Greenspan J, Wolfson M, Keklikian E, Shaffer T, Allen J.
Effects of inspiratory resistive loading on chest wall motion and ventilation: differences between preterm and full-term infants. Pediatr
Res 1992;32:589594.
87. Allen J, Wolfson M, McDowell K, Shaffer T. Thoracoabdominal asynchrony in infants with airflow obstruction. Am Rev Respir Dis 1990;
141:337342.
88. Allen J, Greenspan J, Deoras K, Keklikian E, Wolfson M, Shaffer T.
Interaction between chest wall and lung mechanics in normal infants
and infants with bronchopulmonary dysplasia. Pediatr Pulmonol
1991;11:3743.
89. Sivan Y, Deakers T, Newth C. Thoracoabdominal asynchrony in acute
upper airway obstruction in small children. Am Rev Respir Dis 1990;
142:540544.
90. Benameur M, Goldman M, Eccoffey C, Gaultier C. Ventilation and
thoracoabdominal asynchrony during halothane anesthesia in infants. J Appl Physiol 1993;74:15911596.
91. Goldman M, Pagani M, Trang A, Praud J, Sartene R, Gaultier C. Asynchronous chest wall movement during non-rapid eye movement and
rapid eye movement sleep in children with bronchopulmonary dysplasia. Am Rev Respir Dis 1993;147:11751184.
92. Colin A, Hunter J, Stark A, Wohl M. Normal infants and children have
minimal thoraco-abdominal asynchrony assessed by respiratory inductive plethysmography. Am Rev Respir Dis 1993;147:A966.
93. Sackner M, Gonzalez H, Rodriguez M, Belsito A, Sackner D, Grenvik
S. Assessment of asynchronous and paradoxic motion between the
rib cage and abdomen in normal subjects and in patients with chronic
obstructive pulmonary disease. Am Rev Respir Dis 1984;130:588593.
609

94. Drinker P, Shaw L. An apparatus for the prolonged administration of
artificial respiration. J Clin Invest 1929;7:229247.
95. Diaz C, Deoras K, Allen J. Chest wall motion before and during mechanical ventilation in children with neuromuscular disease. Pediatr
Pulmonol 1993;16:8995.
96. Weiner P, Suo J, Fernandez E, Cherniack R. The effect of hyperinflation on respiratory muscle strength and efficiency in healthy subjects
and patients with asthma. Am Rev Respir Dis 1990;141:15011505.
97. Stick S, Ellis E, LeSouef P, Sly P. Validation of respiratory inductance
plethysmography (Respitrace) for the measurement of tidal
breathing parameters in newborns. Pediatr Pulmonol 1992;14:187191.
98. Laing I, Teele R, Stark A. Diaphragmatic movement in newborn infants. J Pediatr 1988;112:638643.
99. Cohen E, Mier A, Heywood P, Murphy K, Boultbee J. Diaphragmatic
movement in hemiplegic patients measured by ultrasonography.
Thorax 1994;49:890895.
100. Froese A, Bryan A. Effect of anaesthesia and paralysis on diaphragmatic mechanism in man. Anesthesiology 1974;41:242255.
101. Harris R, Giovannetti M, Kim B. Normal ventilatory movement of the
right hemidiaphragm studied by ultrasound and pneumotachygraphy. Radiology 1983;146:141144.
102. Ross-Russell R, Helps B, Elliot M, Helms P. Phrenic nerve stimulation
at the bedside in children: equipment and validation. Eur Respir J
1993;6:13321335.
103. Raimbault J. Technique et rsultats de lexploration lectromyographique
104.
105.
106.
107.
108.
109.
110.
du diaphragme chez le nourrisson et le jeune enfant. Rev Electroencephalogr Neurophysiol Clin 1983;13:306311.
Similowski T, Fleury B, Launois S, Cathala H, Bouche P, Derenne J-P.
Stimulation magnetique cervicale (SMC). Rev Mal Respir 1988;5:609614.
Rafferty GF, Greenough A, Dimitriou G, Polkey MI, Long A, Davenport
M, Moxham J. Assessment of neonatal diaphragmatic paralysis using
magnetic phrenic nerve stimulation. Pediatr Pulmonol 1999;27:224226
Remmers J, Launois S, Feroah T, Whitelaw W. Mechanics of the pharynx in patients with obstructive sleep apnea. Prog Clin Biol Res 1991;
345:261268.
Smith T, Baska R, Francisco C, McCray G, Kunz S. Sleep apnea syndrome: diagnosis of upper airway obstruction by fluoroscopy. J Pediatr 1978;93:891892.
Schwab R, Gupta K, Gefter W, Metzger L, Hoffman E, Pack A. Upper
airway and soft tissue anatomy in normal subjects and patients with
sleep-disordered breathing: significance of the lateral pharyngeal
walls. Am J Respir Crit Care Med 1995;152:16731689.
Condos R, Norman R, Krishnasamy I, Peduzzi N, Goldring R, Rapoport D. Flow limitation as a noninvasive assessment of residual upper-airway resistance during continuous positive airway pressure
of obstructive sleep apnea. Am J Respir Crit Care Med 1994;150:
475480.
Wilson S, Thach B, Brouillette R, Abu-Osba Y. Upper airway patency
in human infant: influence of airway pressure and posture. J Appl
Physiol 1980;48:500504.
10. Assessment of Respiratory Muscle Function in the

Intensive Care Unit
Few areas exist in clinical medicine where problems with the
respiratory muscles play a greater role than in patients managed in an intensive care unit (ICU). The inability of the respiratory muscles to sustain spontaneous ventilation is the primary indication for the main therapeutic modality of an ICU,
namely mechanical ventilation (1). Likewise, respiratory muscle performance is the major issue in deciding the timing and
pace with which mechanical ventilation can be discontinued
(2, 3). Many tests of respiratory muscle function require a high
level of motivation and cooperation, making it much more difficult to obtain reliable measurements in critically ill patients
than in an ambulatory-care setting. This section covers the
tests of demonstrable, or potential, value that directly, or indirectly, relate to respiratory muscle function in the ICU setting.
Consequently, the tests focus predominantly on a patients
need for mechanical ventilation and how this therapeutic modality can be best delivered.
BREATHING PATTERN
Abnormalities in respiratory frequency (fR) and tidal volume
(VT) are extremely common in critically ill patients and, among
other causes, can reflect respiratory muscle dysfunction. In
several studies, an elevated fR has been shown to predict an
adverse outcome in general populations of critically ill patients. In a case-controlled study of patients discharged from
an ICU, fR (p 0.0002) and hematocrit (p 0.01) were the
only continuous variables that predicted readmission to the
ICU (4). Patients readmitted to the ICU had a much higher
mortality than the control patients (42 and 7%, respectively).
In a study of patients who had undergone a cardiopulmonary
arrest, 53% had a documented deterioration in respiratory
function in the 8 hours preceding the arrest (5). Respiratory
frequency was elevated in the majority of these patients
(mean SE, 29 1 breaths/minute), while other routine laboratory tests, including electrocardiograms, showed no consistent abnormalities. These and other studies (6, 7) demonstrate
that tachypnea is an extremely sensitive marker of a worsening clinical status. However, it is also extremely nonspecific,
and a physician needs to undertake additional diagnostic testing to elucidate the nature of the underlying disorder.
Tidal volume (VT), breathing frequency (fR), and minute
ventilation are easy to measure in intubated patients, and the
values are continuously displayed on virtually all modern mechanical ventilators. However, the accuracy and reliability of
the instrumentation used for volume measurements have undergone remarkably little evaluation (8). To ensure reliability
of volume measurements, a simple handheld spirometer is
preferred. Tidal volume is rarely measured in the nonintubated patient, because these patients have a poor tolerance of
mouthpieces. Moreover, such instrumentation usually causes
a spurious increase in VT and decrease in frequency (9, 10).
Systematic studies have not been undertaken in large populations of healthy subjects to define the normal range for
breath components. Moreover, the values depend on whether
recordings are made nonobtrusively or with instrumentation
requiring the use of a mouthpiece (9, 10). The largest study using nonobtrusive methodology in healthy subjects (n 65) reAm J Respir Crit Care Med Vol 165. pp 610623, 2002
vealed
a mean SD VT of 383 91 ml, fR of 16 2.8 breaths/
DOI: 10.1164/rccm.AT10-02
Internet address:
www.atsjournals.org
minute,
and minute
ventilation of 6.01 1.39 L/minute (11).
Additional data in smaller study populations are consistent

with these values. Breath components display considerable
breath-to-breath variability, and thus it is important to base
the measurement on an adequate sampling period; unfortunately, investigations have not been conducted to define the
optimal sampling duration. Calculations of breath components
based on the mean of approximately 250 breaths revealed
good reproducibility on a day-to-day basis in healthy subjects
(coefficient of variation less than 9%) (12).
Rapid shallow breathing is particularly common in critically ill patients (Figure 1) (13), and some (1416), but not all
(17, 18), investigators have related its development to respiratory muscle fatigue. Even when rapid shallow breathing has
been demonstrated following the induction of fatigue in
healthy volunteers (16), it occurred when the subjects were
employing only a small fraction of the pressure-generating capacity of the inspiratory muscles, making it unlikely that this
alteration in breathing is a manifestation of fatigue per se.
Critically ill patients are susceptible to many simultaneous
challenges, and rapid shallow breathing in these patients may
be the result of several mechanisms, including increased mechanical load, chemoreceptor stimulation, operating lung volume, reflexes originating in the lungs and respiratory muscles,
altered respiratory motoneuron discharge patterns, sense of
effort, and cortical influence. If rapid shallow breathing were
an effective strategy for avoiding respiratory muscle fatigue in
critically ill patients, one would expect a negative correlation
between the degree of rapid shallow breathing and a measure
of respiratory muscle fatigue. In 17 patients who failed a trial
of weaning from mechanical ventilation, Jubran and Tobin
(18) found no relationship (r 0.08) between frequency-totidal volume ratio (fR/VT) and tensiontime index. Vassilakopoulos and coworkers (19) have confirmed the lack of a relationship between fR/VT and tensiontime index (r 0.16) in
30 patients who failed a weaning trial. In the latter study (19),
tensiontime index decreased from a value of 0.162 0.032
(SD) in patients who could not be weaned to 0.102 0.023 at
the time of weaning success. When a large number of variables
reflecting respiratory muscle function, lung volumes, and mechanics were entered into a logistic regression model, fR/VT
and tensiontime were the only variables that were significantly related to weaning outcome.
While the pathophysiological mechanism(s) responsible for
rapid shallow breathing is unknown, this does not detract from
its use in clinical decision-making, especially with regard to
the timing and pace of the weaning process (see subsequent
text). Moreover, initial studies of computerized closed-loop
adjustments of ventilator settings, based on VT and fR, with or
without end-tidal PCO2 and pulse oximetry, provide encouraging data suggesting that ventilator adjustments based on changes
in breathing pattern might help in expediting the weaning process (2022).
Asynchronous and paradoxic motion of the rib cage and
abdomen is commonly detectable by inspection and palpation
in critically ill patients. The magnitude of abnormal motion can
be quantified with magnetometers or inductive plethysmography
combined with the KonnoMead method of analysis (23) (see
ESTIMATION OF VENTILATION BASED ON CHEST WALL MOTION:
KONNOMEAD DIAGRAM in Section 6 of this Statement). Ashutosh, Gilbert, and coworkers (24, 25) showed that patients dis-
611
playing asynchronous rib cageabdominal motion had an increased risk of ventilatory failure necessitating mechanical
ventilation (24) and a poor prognosis (24). Subsequently, abdominal paradox and respiratory alternans, i.e., cyclic alteration in the relative contribution of the rib cage and diaphragmatic muscle groups, were thought to reflect respiratory muscle
fatigue (14, 26, 27). Such an interpretation has major implications for the critically ill patient being weaned from mechanical ventilation. Because muscle rest with mechanical ventilation is the main means of reversing fatigue, the presence of
paradox would prohibit the discontinuation of mechanical
ventilation. Patients who fail a weaning trial, as a group, exhibit greater abdominal paradox than successfully weaned patients (28, 29), but there is considerable overlap among individual patients of the two groups. This lack of discrimination
between the two groups of patients partly stems from the fact
that the more seriously ill patients also switch between predominant use of the rib cage and diaphragmatic muscles (14,
28, 29). Moreover, in systematic experimental studies, respiratory muscle fatigue was shown to be neither necessary nor sufficient for the development of abnormal rib cageabdominal
motion (30, 31). Thus, quantification of abdominal paradox
alone is not helpful in detecting respiratory muscle fatigue or
predicting the development of respiratory failure. However,
studies in small numbers of patients suggest that a global measure of overall asynchronous and paradoxic motion of both
the rib cage and abdomen might be useful in predicting ventilatory failure (14, 24, 25, 2830). Because quantification of abnormal rib cageabdominal motion is relatively complex, such an
index cannot be recommended for general clinical use without
undertaking controlled prospective trials to determine if such
a measurement is superior to simpler tests that are more easily
performed (32).
LUNG VOLUMES
Few studies have been conducted in critically ill patients examining the usefulness of lung volume measurements. In a
study of 10 patients with the GuillainBarr syndrome, Chevrolet and Deleamont (33) found that monitoring vital capacity
three times a day was helpful in predicting the need for ventilator support. Five patients required intubation approximately
12 days after the onset of neurological symptoms; vital capacity was 0.82 L (range, 0.651.00 L) immediately before intubation in these patients. In contrast, mean SD vital capacity
was 2.40 0.57 L in another five patients who were managed
without intubation. In general, when vital capacity fell by 50%
Figure 1. A breath-by-breath
plot of respiratory frequency
and tidal volume in a patient
who failed a weaning trial.
The arrow indicates the point
of resuming spontaneous
breathing following discontinuation of ventilator support.
Rapid, shallow breathing developed almost immediately,
suggesting the prompt establishment of a new steady
state. Although it has been
considered that rapid, shallow
breathing may reflect the
presence of respiratory muscle
fatigue, its almost instantaneous development without subsequent
progression is difficult to reconcile with the development of respiratory
muscle fatigue. Reprinted by permission from Reference 13.
from the first recorded value, ventilator support became necessary within the next 36 hours.
Rieder and coworkers (34) examined the accuracy of vital
capacity in predicting the need for mechanical ventilation in
patients with myasthenia gravis. The study consisted of a retrospective chart review of five patients who experienced 10
episodes of acute respiratory failure. Repeated measurements
of vital capacity every 4 hours did not help in predicting the
need for intubation and mechanical ventilation, which was instituted in four of the episodes. Indeed, the lowest values of vital
capacity tended to be found in the patients who did not receive mechanical ventilation, mean SD 11 5 ml/kg, compared with 21 9 ml/kg in the patients who received ventilator support. Although the study population is small, the results
suggest that vital capacity is unlikely to be useful in this disease. The inability to predict deterioration may be due to the
erratic course of this disease, which involves sudden deterioration (see STATIC LUNG VOLUMES in Section 1 and DEVICES USED
TO MONITOR BREATHING in Section 6 of this Statement).
PRESSURE MEASUREMENTS
The respiratory system is an elastic structure that requires the
generation of force (pressure) for its displacement (35). In
many critically ill patients, positive pressure is delivered to the
lungs through an endotracheal tube by a mechanical ventilator
with the aim of improving arterial blood gases and unloading
the respiratory muscles (3638). The total respiratory system
is composed of complex structures, namely the lungs, the upper and lower rib cage, and the diaphragmabdominal compartments, each displaying different mechanical properties; in
disease states, the situation is even more complicated due to
time-constant inhomogeneity, expiratory flow limitation, etc.
(see ASSESSMENT OF THE FUNCTION OF THE ACTIVE CHEST WALL
in Section 6 of this Statement). Nevertheless, assessment of
respiratory system mechanics is commonly based on a relatively simple equation of motion (Equation 1) (35):
(1)
Pappl = 1 C V + RV
where Pappl is the pressure applied to the system by either the
ventilator or the combined action of the ventilator and the pa
tients inspiratory muscles, V and V are the inflation volume
and flow, respectively, C is the respiratory compliance, and R
is the flow resistance. The endotracheal tubes and the ventilator circuits are rather stiff structures, such that a patients
compliance provides a good representation of the overall elastic properties of the patientventilator ensemble. In contrast,
the endotracheal tube poses a substantial, highly flow-dependent resistance, which is important from the standpoint of a
patients respiratory muscle activity, because this resistance
has to be overcome during lung inflation (39, 40) and it may
retard expiratory flows, thus promoting dynamic hyperinflation (41, 42). Due to the flow characteristics of the endotracheal tubes, Equation 1 becomes Equation 2:
2
Pappl = 1 C V + Rt + ( k V + k V )
(2)
1
where Rt is the nonohmic component of total flow resistance,

and k1 and k2 are constants related to laminar and turbulent
flow, respectively. The additional work of breathing due to the
endotracheal tube can be considerable at high levels of minute
ventilation, and it may be negligible at low inspiratory flows.
Inspiratory effort is usually increased in critically ill patients because of abnormal respiratory mechanics, i.e., low
compliance and high flow resistance (4347). During mechanical ventilation, a variable portion of the ventilatory workload
is decreased such that a patients inspiratory muscles are un-
612
loaded by an amount that should be proportional to the degree of mechanical support (4856). However, in some instances, a patients inspiratory effort during ventilator-assisted
breaths differs only slightly from that during unassisted breathing due to several factors: excessive ventilatory drive consequent to either metabolic factors (e.g., sepsis, fever, etc.) or
psychologic (e.g., pain, anxiety, etc.) phenomena; substantial
time lag between the onset of a patients inspiratory effort and
full machine support due to delayed opening of the ventilator
circuit valves; ventilator-inspiratory flow that does not meet
patient demands, especially at the onset of inspiration; intrinsic positive end-expiratory pressure (PEEP); and excessive VT
that requires a long expiratory time, during which ineffective
patients efforts may occur. If a patients effort remains significant during mechanical ventilation, the respiratory muscles
cannot recover from fatigue. Accordingly, progressive reduction and eventual discontinuation of mechanical ventilation
can be very difficult (2, 3, 57).
Measurement of pressure at the airway opening is easy in
ventilator-dependent patients because a stiff endotracheal tube
bypasses the compliant upper airway, allowing the rapid transmission of changes in alveolar pressure to the airway opening
even in presence of time-constant inhomogeneity (58, 59).
Airway Pressure Contour
Consistent with Equation 1, simple inspection of the airway

pressure contour in a ventilator-dependent patient can provide
useful information on the activity of a patients respiratory
muscles. With a given inspiratory flow waveform, any deviation
from the relaxed configuration indicates active contraction of a
patients inspiratory muscles (Figure 2). The work performed
by the patient can be quantified as the difference in area between the actual versus the relaxed airway pressure contour
plotted against either time or volume (48, 49, 53). A patient actively contributes to the total work of breathing (WOB), either
intentionally or unintentionally, whenever his or her ventilatory
demands exceed the inspiratory flow rate setting on the ventilatora phenomenon called flow deprivation. Some modern
ventilators have the capability of detecting such an occurrence
with their operational software and can compensate for it.
Maximum Inspiratory Pressure
Measurement of maximum inspiratory pressure (PI,max) has

long been used in the clinical setting to assess inspiratory mus-
VOL 166
2002
cle strength (see Section 2 of this Statement). The maneuver

consists of a maximum inspiratory effort against a closed airway and requires a considerable degree of patient cooperation
and coordination (60). Although ventilator-dependent patients
may display poor cooperation in the execution of voluntary
maneuvers, even in these patients values of PI,max have been
used to predict successful weaning (61). Marini and coworkers
(62) suggested an approach for the standardization of the
measurement of PI,max in such patients. They used a unidirectional valve to permit exhalation while inhalation was blocked,
thereby allowing patients to perform the maximal inspiratory
effort at a lung volume approaching residual volume, where
PI,max is expected to be maximal; the highest PI,max values
were generally reached after 1520 efforts or after 1520 seconds of airway occlusion (62). However, subsequent work has
shown that even employing such a standardized approach, the
reproducibility of PI,max values in ventilator-dependent patients is poor. True PI,max in ICU patients is often significantly underestimated, being both patient- and investigatordependent; even highly reproducible PI,max measurements at
any one sitting do not reliably reflect maximal efforts (63).
Although a high value of PI,max, together with other measurements, may indicate that mechanical ventilation can be
discontinued, a low PI,max value may reflect a submaximal effort due to poor patient coordination and cooperation. In addition, the problem of insufficient reproducibility may limit
the clinical usefulness of this test, at least when used in isolation.
Pressure at the Airway Opening in the First 0.1 Second
The decrease in airway pressure at 0.1 second (P0.1) after

commencement of a tidal inspiratory effort against an occluded airway has commonly been used as an index of neuromuscular ventilatory drive (6466). The value of P0.1 depends
not only on the inspiratory center output, but also on several
other factors, including: (1) the intact neural pathway to the
inspiratory muscles; (2) electro-mechanical coupling; and (3)
the pressure-generating capacity of the inspiratory muscles
and their velocity of shortening (see TESTS OF RESPIRATORY
CONTROL in Section 1 of this Statement).
Although a high value of P0.1 always indicates enhanced
respiratory center activity, a low value may signify not only reduced center output, but also deterioration in the neural pathway, impaired electro-mechanical coupling, and impaired pres-
Figure 2. Subtraction of the area subtended by the inflation pressure (Prs)volume (right) or -time (left) curve in
the presence of inspiratory muscle activity (B) from
that recorded during passive inflation (CMV, controlled
mechanical ventilation; A) yields the work of inspiration
performed by the patient (AMV, assisted mechanical
ventilation; shaded area, C) and the pressuretime
product ( P dt) of the inspiratory muscles. Reprinted
sure-generating capacity due to respiratory muscle weakness

and fatigue. Furthermore, in the presence of dynamic hyperinflation and PEEPi, P0.1 measured as subatmospheric pressure
100 seconds after the apparent onset of effort, neglects the respiratory effort required to draw down PEEPi. These factors
must be kept in mind when evaluating P0.1 in the ventilatordependent patient. In a patient with abnormal airway resistance breathing through an intact upper airway, the true value
of P0.1 may be underestimated because of the long time constant, retarding the transmission of pressure changes from the
alveoli to the mouth (67). In contrast, in a ventilator-dependent patient who has a rigid endotracheal tube bypassing the
compliant region of the upper airway, P0.1 may more accurately reflect esophageal pressure changes and thus neuromuscular drive (68).
P0.1 is generally increased in patients with acute respiratory failure (69), although an authoritative study on the reproducibility of P0.1 measurement in the ICU setting has yet to
be performed. In mechanically ventilated patients, P0.1 has
been used to: (1) assess respiratory center output while changing the fraction of inspired oxygen (FIO2) (70); (2) predict
weaning outcome values of P0.1 6 cm H2O suggesting that
discontinuation of mechanical ventilation is likely to be unsuccessful, while values 4 cm H2O suggest that weaning is likely
to be successful (71, 72); and (3) set the appropriate level of
assisted ventilation (73) and pressure support (56).
Measurement of P0.1 does not require sophisticated equipment, and a two-way valve and differential pressure transducer are sufficient. In ventilator-dependent patients, the procedure can become even simpler because the valves necessary
for the measurement already exist in the ventilator. Indeed, to
trigger a mechanical breath, the patient has to generate a
small negative pressure at the airway opening, which opens
the valves and allows the passage of inspiratory gas. In some
ventilators, e.g., the Siemens Servo 900C, the time to open the
valve exceeds 100 milliseconds, thus providing sufficient time
to measure the P0.1. This modification has been used successfully by Fernandez and coworkers (69) and Conti and coworkers (74) to measure P0.1 in patients with intrinsic PEEP (see
subsequent text). Kuhlen and coworkers (75) introduced an
automated method to measure P0.1 in ventilator-dependent
patients using a standard ventilator (i.e., the Drger Evita).
These authors correctly underscored the breath-by-breath variability of the measurement. In this connection, it has to be
mentioned that expiratory muscle activity may increase the
P0.1 variability and generate unreliable values.
In summary, P0.1 can be measured in ventilator-dependent
patients and taken as an index of neuromuscular drive. High values of P0.1 are considered to predict weaning failure, whereas
low values of P0.1 suggest that mechanical ventilation can be discontinued, provided that factors that can impair the pressuregenerating capacity of the inspiratory muscles are excluded.
Pleural Pressure and Derived Variables
Recording of changes in pleural pressure (Ppl) represents a basic means of quantifying respiratory muscle activity. Recording
the tidal swings in esophageal pressure (Pes) (as an estimate of
Ppl) and a knowledge of the passive properties of the chest wall
over lung volume allow the quantification of the pressure developed by the muscles, also referred to as muscular pressure
(Pmus). This measurement has facilitated the gathering of important information in the field of mechanical ventilation, both
in quantification of the effort performed by patients receiving
ventilator assistance and for comparison of different modes of
assisted ventilation (see ESOPHAGEAL, GASTRIC, AND TRANSDIAPHRAGMATIC PRESSURES in Section 2 of this Statement).
613
Two main variables can be derived: work of breathing

(WOB) and pressuretime product (PTP) (see ASSESSMENT OF
THE FUNCTION OF THE ACTIVE CHEST WALL: CAMPBELL DIAGRAM in Section 6 and PRESSURETIME PRODUCT in Section 4 of
this Statement) (76). For both calculations, knowledge of the
compliance curve of the chest wall is necessary; several methods have been used for this measurement in mechanically ventilated patients. Some investigators employ an assumed chest
wall compliance value, amounting to 4% of the predicted vital
capacity per cm H2O (47, 77). Assuming that the chest wall
compliance is linear over the range of lung volumes studied, a
straight line can be traced over lung volume to delineate the
pressure reference. This method has the major disadvantage
of not measuring the true chest wall compliance, which has
been shown to be substantially modified in some intubated patients with acute respiratory failure (43, 44, 78). A preferable
approach is to measure the pressurevolume (PV) relationship, using Pes recordings over the range of lung volumes
studied, during passive inflation under controlled mechanical
ventilation. This makes it necessary to heavily sedate the patient and/or to abolish spontaneous activity by hyperventilating the patient. Again, for simplicity, this curve can be assumed to be linear, or it can also be superimposed graphically
over each studied breath, assuming that the end-expiratory
lung volume can be estimated (79). The latter approach allows
quantification of active expiratory effort.
WOB is measured from pressurevolume loops using the
Campbell diagram method (80) (see ASSESSMENT OF THE FUNCTION OF THE ACTIVE CHEST WALL: CAMPBELL DIAGRAM in Section
6 of this Statement). The measurement of the WOB in ICU patients raises no unique concerns, although recognition of some
limitations of the measurement has led to the use of alternative
methods of measuring respiratory energy expenditure. Measurement of WOB does not allow quantification of isometric efforts,
such as an effort against a closed airway or inefficient or
wasted inspiratory efforts that fail to trigger a ventilator (81
83); in addition, measurement of WOB is probably inadequate
for the comparison of settings where part of the change in VT is
achieved by the ventilator. In such a case, an increase in WOB
may be recorded in the absence of a true change in muscular effort. Such concerns about WOB have been recognized for a long
time (76) and have prompted the use of other indices to quantify
respiratory muscle effort in the ICU, especially PTP (51, 76).
Calculation of the PTP necessitates the integration of the
area under the Ppl curve versus time. It also requires, however, a solution to the problem of determining the beginning
of the inspiratory effort and of using PEEPi to reference the
true beginning of inspiration to the chest wall relaxation line
(52). Different methods have been proposed to perform this
calculation (which is applicable as well to the transdiaphragmatic pressure [Pdi]), and the termination of inspiratory effort
has been based on the inspiratoryexpiratory flow transition
point (52) or once the pressure has returned back to its baseline level (84). PTP can be expressed on a per-breath or a perminute basis; today, it is probably one of the most useful tools
for quantifying respiratory muscle effort in mechanically ventilated patients (18, 52, 80, 8593) (see PRESSURETIME INDEX
OF INSPIRATORY MUSCLES in Section 5 and PRESSURETIME
PRODUCT in Section 4 of this Statement). PTP can be separated
into a number of components: the effort made to counterbalance intrinsic PEEP; the effort to trigger the ventilator; and
the effort to inflate the chest in the posttrigger phase (83, 87,
93) (Figure 3). However, the correct interpretation of this
measurement when end-expiratory lung volume (FRC) undergoes concurrent change remains a question of controversy
that has yet to be solved and deserves further investigation.
614
The presence of a phase lag between the onset of the decline in inspiratory pressure decay and the time at which the
flow reaches zero indicates the presence of positive alveolar
pressure at end-expiration. Because this positive pressure can
result from either elastic recoil pressure generated by hyperinflation or from activity of abdominal and/or thoracic expiratory muscles, interpretation of this pressure is extremely difficult. Different methods have been proposed to solve this
problem, using the Pdi (94) or the expiratory swings in gastric
pressure (Pga) (95) to estimate the part related to expiratory
muscle activity (see subsequent text). When Pga is not available, one can calculate a range of values between two extremes or bounds: at one extreme, it is assumed that all of the
pressure drop before reaching zero flow is related to dynamic
hyperinflation, and at the other extreme, it is assumed that all
of this pressure drop is related to the relaxation of the expiratory muscles (79).
Measurements of Ppl and derived variables are very helpful in research. In the clinical setting, however, it has gained
relatively little acceptance despite the availability of commercial systems. Some investigators have suggested that Ppl may
help in understanding the reasons for the inability to tolerate
discontinuation from mechanical ventilation (9698). Although
Ppl has potential implications for setting the ventilator, clearcut data are not available on which to base individual targets
for the employment of these measurements. In addition, a
number of artifacts, including interference by cardiac contractions, make automatic calculations unreliable, especially regarding the detection of the end and the onset of both inspiratory and expiratory efforts.
Intrinsic PEEP
In ventilator-dependent patients, end-expiratory alveolar pressure may remain positive even in the absence of external
PEEP whenever the time required to decompress the lungs to
the elastic equilibrium volume is shorter than the expiratory
time (tE) available before the next inspiration (99, 100). This
end-expiratory elastic recoil pressure (Pel,rs) has been termed
occult PEEP, auto PEEP (101), and PEEPi (102). Both abnormalities in a patients respiratory mechanics and inappropriate
ventilator settings may cause PEEPi (100, 101). As in the case
of set PEEP, PEEPi may decrease cardiac output (103, 104),
and by thus impairing respiratory muscle perfusion, contribute
to respiratory muscle dysfunction. Furthermore, PEEPi poses
an inspiratory pressure threshold load during both spontaneous breathing and patient-triggered, assisted modes of mechanical ventilation. Indeed, PEEPi must be fully counterbalanced by the contraction of a patients inspiratory muscles
before the ventilator can be triggered (99), giving Equation 3:
2
Pappl = PEEPi + V C + Rt + ( k 1 V + k 2 V )
(3)
Therefore, during assisted modes of mechanical ventilation,

the total effort needed to trigger the ventilator does not correspond to the small negative pressure (i.e., 1 to 2 cm H2O)
usually set. Clearly, high levels of PEEPi, such as those in patients with an acute exacerbation of chronic obstructive pulmonary disease (COPD) (94, 105, 106) or asthma (107), can
increase significantly the magnitude of a patients effort to
trigger the ventilator. In some instances, the achieved decrement in Ppl may be smaller than the level of PEEPi, resulting
in failure to trigger the ventilator and the occurrence of ineffective or wasted efforts (8183, 108111). Application of a
low level of PEEP and/or continuous positive airway pressure
(CPAP) can improve patientventilator interaction and reduce significantly the magnitude of inspiratory effort during
VOL 166
2002
Figure 3. Illustration of method

used to calculate the pressure
time product (PTP) for the
esophageal pressure (Pes) and
to partition it into its constituent components: efforts required to overcome PEEPi
(PTPpeepi), triggering effort
(PTPttr), and posttriggering
effort (PTPpost). PTPpeepi constitutes a load imposed during the whole inspiration,
whereas the triggering effort
represents only a short delay
before being transformed into
volume. CW chest wall; E
expiration; Pmus muscular pressure; P PEEPi; I
onset of inspiratory flow. Reprinted by permission from
Reference 93.
assisted ventilation and weaning by counterbalancing PEEPi

at least in part (91, 93, 111).
Intrinsic positive end-expiratory pressure can be measured
in two ways: by the end-expiratory airway occlusion technique
and by simultaneous recording of flow and Ppl.
End-expiratory airway occlusion technique. End-expiratory
airway occlusion (EEO) for measurement of static PEEPi can
be performed manually at the expiratory port of the ventilator
during the last 0.5 seconds of the expiration (112). Alternatively, a pneumatic valve can be used for rapid occlusion at the
airway opening (111, 113, 114). This second technique has the
advantage of avoiding problems related to gas compression in
the ventilator tubing, although some variability can occur due
to timing of the occlusion. However, near the end of expiration, the volume expelled per unit of time is very small because of extreme airway compression, and therefore, small
changes in tE are not expected to affect substantially the magnitude of PEEPi, at least for a tE of around 3 seconds (111).
Measurement of PEEPi can be easily performed with ventilators that are already equipped with an end-expiratory occlusion hold option, either a manual operating button (e.g., the
Siemens Servo 900C or 300) (115118) or a hardware/software
facility (e.g., the enhanced 7200 series of the Puritan-Bennett
ventilator), for rapid occlusion of the expiratory port exactly
at the end of a tidal expiration. With ventilators that do not include this option, airway occlusion can be performed using
manually operated valves (119, 120).
Simultaneous recording of flow and Ppl. The accurate measurement of PEEPi can pose a problem in patients with spontaneous activity of the inspiratory muscles, although some authors have been able to relax patients and obtain an apparent
end-expiratory plateau (91, 112). However, in actively breathing patients, measurement of PEEPi might require the insertion of an esophageal balloon to estimate changes in Ppl. Although the occlusion test has been suggested for validation
of the esophageal balloon technique (121), its use in supine
critically ill patients has been questioned and a legitimate
study has never been really performed. As illustrated in Figure 4, PEEPi is measured as the decrease in Ppl preceding the
onset of inspiratory flow. PEEPi measured in this way represents dynamic PEEPi (PEEPi,dyn), which is significantly less
than static PEEPi (122). This method of measuring PEEPi,dyn
is valid provided that the expiratory muscles are relaxed at the
end of the expiration. Indeed, if the expiratory muscles are actively contracting during expiration, the decrease in Ppl at
end-expiration could be due to expiratory muscle relaxation
615
rather than inspiratory muscle contraction (123). This may

also occur in ventilator-dependent patients (94, 95).
To obtain the correct value of PEEPi, i.e., without positive
end-expiratory alveolar pressure due to expiratory muscle activity, a number of approaches have been employed. Appendini and coworkers (94) suggested subtracting the corresponding change in Pga from the change in Ppl during the time
interval from the onset of the inspiratory effort, as indicated
by the swing in Pdi, to the start of inspiratory flow (Figure 4).
Lessard and coworkers (95) proposed that the expiratory rise
in gastric pressure should be subtracted, assuming a complete
transmission of pressures generated by all or most of the expiratory muscle activities to the abdominal cavity. Recently,
Yan and coworkers (124) used the Campbell diagram method
to measure PEEPi,dyn in spontaneously breathing subjects
and to correct for expiratory muscle activity without requiring
the insertion of a gastric balloon. In this connection, it has to
be mentioned that some nasogastric tubes used for nutrition
include a balloon-catheter system for measurement of changes
in Ppl, thus limiting the invasiveness of the maneuver (125).
Yan and coworkers (124) urged caution when interpreting the
measurement of PEEPi,dyn in actively breathing subjects, because of the confounding effects of the viscoelastic properties
of the respiratory system, postinspiratory inspiratory muscle
activity, and expiratory muscle activity. Moreover, it has to be
considered that during the interval that PEEPi is being measured, contraction of the diaphragm can increase Pga while
concomitant relaxation of the expiratory muscles can decrease
Pga such that the relative magnitude of those two separate
activities remains unpredictable. Furthermore, expiratory rib
cage muscles can be active and also affect Pes. Therefore,
even after using Pga to correct the recorded PEEPi,dyn, the
corrected value may not truly quantify the magnitude of dynamic hyperinflation (124).
In summary, due to its important clinical implications,
PEEPi should be measured in ventilator-dependent patients.
The end-expiratory airway occlusion remains the reference
maneuver, although interpretation is extremely difficult in the
case of expiratory muscle recruitment (95).
Work of Breathing
Mechanical work (W) implies that an applied force or pressure (e.g., Pappl in Equation 1) is producing some displacement in a system, e.g., change in volume (V), according to
Equation 4:
w = Pappl V =
Papp V
(4)
that represents the area subtended by the PV curve. WOB can

be expressed in power (per unit of time) or per liter of ventilation (J/L). The relationship between work and minute volume,
however, is usually nonlinear. Because the reduction of excessive patient WOB due to abnormal respiratory mechanics is a
major goal of mechanical ventilation, measurement of WOB is
of substantial interest in ventilator-dependent patients. However, the technique can be invasive and is not adequately standardized.
During controlled mechanical ventilation (CMV), the passive work is measured according to Equation 4. This measurement has been used not only to gather information on a patients respiratory mechanics, but also to estimate a patients
WOB by mimicking a spontaneous breathing pattern (126).
The use of such measurement to predict whether a ventilatory
workload is compatible with unsupported breathing is too
complex for routine use in weaning management (127129).
During assist-control ventilation, as suggested by Marini and
Figure 4. Experimental
record illustrating the
method used to determine the level of dynamic
intrinsic PEEP (PEEPi,dyn)
during spontaneous unoccluded breathing efforts in a representative
patient with active expiratory muscles. From
top to bottom, tracings
represent flow ( V), pleural pressure (Ppl), gastric
pressure (Pga), and transdiaphragmatic pressure
(Pdi). The first vertical line
indicates the point corresponding to the onset of the inspiratory effort (Pdi swing). The
second vertical line indicates the point corresponding to the start of
the inspiratory flow. The
dotted horizontal line represents zero flow. in inspiratory flow; ex
expiratory flow. (A) Tidal volume 0.46 L. Note that expiratory flow
ends abruptly before inspiration, whereas the Pdi and Ppl swings have
already begun and Pga has remained constant during that interval. In
this case PEEPi,dyn was measured as the negative deflection in Ppl between the point corresponding to the onset of the Pdi swing and the
point of zero flow. (B) Tidal volume 0.33 L. Note that Pga increases
throughout most of the expiration. By contrast, Pga becomes less positive from the onset of the inspiratory effort indicated by the start of
positive Pdi swing to the start of inspiratory flow. In this case PEEPi,dyn
was measured as the negative Ppl deflection between the point corresponding to the onset of the Pdi swing and the point of zero flow subtracted by the amount of Pga negative deflection observed in that interval. Reprinted by permission from Reference 94.
coworkers (49, 50) and Ward and coworkers (53), the patients
WOB can be estimated noninvasively by comparing the pressurevolume (or pressuretime) relationship with that during
CMV. Differences in WOB among patients, even in the same
patient under different circumstances, may be used to compare the effects of the ventilatory settings, particularly volume
and flow, the internal diameter of the endotracheal tube, the
ventilator devices, or for evaluation of a patients response to
therapy, such as bronchodilators (93, 130132). During assisted modes of ventilation, a patients WOB can be measured
with the esophageal balloon-catheter technique using the
Campbell diagram (Figure 5), which also allows partitioning
between elastic and resistive work, and the detection of expiratory muscle activity. The presence of a positive alveolar
pressure at end-expiration (i.e., intrinsic PEEPi) can considerably increase the area enclosed in the loop and involve supplemental elastic work. Measurement of WOB may help in deciding an appropriate level or type of ventilator assistance and
avoid both excessive and insufficient support. For instance,
continuous flow systems seem to offer less resistance to
breathing than demand flow (i.e., pressure-triggered) systems
and can decrease a patients WOB (86, 93, 133135) (see ASSESSMENT OF THE FUNCTION OF THE ACTIVE CHEST WALL: CAMPBELL DIAGRAM in Section 6 of this Statement).
There is little doubt that measurement of WOB could be
useful in many clinical conditions in ventilator-dependent patients. Automatic systems for its measurement are now available (99, 136138). However, the superiority of the measurement of WOB, over more simple monitoring, for example
over P0.1 (56, 74) or breathing pattern (13), has not been
proved (76). The esophageal balloon technique is unlikely to
616
be used routinely for monitoring the magnitude of inspiratory

effort and WOB. Whether measurements of Ppl could improve useful clinical-decision making and patient management in certain situations, such as the difficult-to-wean patient
or the patient who unexpectedly fights the ventilator (Figure
6), remains to be determined.
Transdiaphragmatic Pressure
Transdiaphragmatic pressure can be measured in ICU patients for research or diagnostic purposes when diaphragmatic
dysfunction is suspected. It requires simultaneous recordings
of Pes and Pga; Pdi is obtained by measuring the differential
pressure between these two signals. Because Pga usually rises,
or is positive, during inspiration whereas the Pes signal is usually negative, subtraction of Pes from Pga is, in effect, the sum
of the two tidal excursions. This measurement is described in
more detail elsewhere. The potential applications of this measurement in ICU patients are summarized in subsequent text
(see TECHNIQUES FOR PRESSURE MEASUREMENTS in Section 2 of
this Statement).
Absolute values of the Pdi swings. Although this measurement has not been used for quantification of respiratory muscle dysfunction, its relation to the amplitude of VT is a reflection of the degree of mechanical coupling between the respiratory
muscles and the respiratory system. While a normal subject
may require a Pdi swing of 5 cm H2O to obtain a VT of 500 ml,
a severely diseased patient, as often seen in the ICU, may develop up to 30 cm H2O to achieve only half of this volume
(85). The same reasoning can be used for Pes swings. Minimizing tidal Pdi has also been proposed to be a target for titration
of pressure support ventilation (139).
Maximum Pdi (Pdi,max) and mean Pdi over Pdi,max. This
ratio has been proposed in the laboratory setting to predict
the risk of respiratory muscle fatigue (140). Although it is of
potential interest in the field of intensive care medicine, its use
is limited by the extreme difficulty of obtaining reliable measurements of Pdi,max in this setting (141). It is necessary to
motivate the patient and, in particular, to display the pressure
traces while the patient performs the maneuver. The same
Figure 5. Loops of esophageal pressure against VT

used for computation of
the work of breathing in
a patient successively ventilated with pressure support (PS) of 0, 10, 15,
and 20 cm H2O. The work
per breath is represented
by the cross-hatched area
subtended by the inspiratory part of the pressure
volume loop on the left
and the relaxation curve
of the chest wall on the
right. Arrows indicate the
beginning and end of inspiration. The relaxation
curve of the chest wall is
passed through the elastic recoil pressure of the
chest at end exhalation,
and the distance between the beginning of
the pressurevolume loop (arrow) and the relaxation curve represents
intrinsic PEEP. The concomitant decrease in the work per breath and in
the respiratory rate (rr) results in a large reduction in the work of
breathing per minute (power). Reprinted by permission from Reference 55.
VOL 166
2002
problem exists for the diaphragmatic tension time index (Pdi/

Pdi,max inspiratory time [tI]/total time of the respiratory cycle [tTOT]), which incorporates the duty cycle (142). Also, calculation of mean Pdi and tI/tTOT is complicated by the difficulty in determining the baseline value of Pdi in such patients
(84) (see MAXIMAL STATIC INSPIRATORY AND EXPIRATORY PRESSURES in Section 2 of this Statement).
The absolute values of Pdi,max can be used to estimate the
strength of the diaphragm, but normal values have a wide
range: varying between 60 and 240 cm H2O. The Pdi value obtained during a sniff maneuver constitutes an attractive alternative because it is much simpler to perform. However, its reliability in intubated patients has not been rigorously tested.
The ratio of the inspiratory Pga swings to Pdi. The Pga/Pdi
ratio has been proposed as a means of assessing the diaphragmatic contribution to the tidal effort (143). The higher this ratio, the greater is the diaphragmatic contribution to the tidal
breathing and the smaller the contribution of other inspiratory
muscles, which all contribute to the negative intrathoracic
pressure swings. In ICU patients, this contribution is often
small (frequently 20%), indicating that recruitment of accessory muscles is a common occurrence (141144). A precise
quantification based on this index is made problematic by the
fact that Pga is influenced by abdominal compliance, which can
vary with abdominal muscle tone, position, and other factors
related to abdominal contents. It is a crude yet useful index,
nevertheless, for the detection of diaphragmatic dysfunction or
paralysis, especially in the postoperative period (144148). A
negative Pga/Pdi ratio indicates severe diaphragmatic (or phrenic
nerve) dysfunction and represents the equivalent of a paradoxical abdominal motion. Correct interpretation of this measurement requires careful avoidance of two pitfalls. First, the incorrect placement of the gastric balloon in the lower esophagus
mimics exactly severe diaphragmatic dysfunction; this problem
is not trivial, because the main criterion for checking the correct placement of the gastric balloon is often the shape of the
signal (144). Second, recruitment of abdominal muscles during
expiration is followed by a sudden relaxation at the beginning
of the next inspiratory effort, which can also mimic diaphragmatic dysfunction (94, 95, 123, 149). Because this is a frequent
occurrence in ICU patients, special attention should be paid to
this problem (and ideally an abdominal wall relaxation curve
should be obtained).
Examination of the Pga signal during expiration. This allows detection of phasic expiratory muscle activity, and it may help in
the differentiation of dynamic hyperinflation from expiratory
muscle recruitment as the mechanism that explains the presence of a phase lag between the beginning of the negative Pes
swings and the onset of inspiratory flow, which is often referred to as intrinsic PEEP (94, 95, 123, 149). Ideally, measurement of pressures should be coupled with an estimate of
thoracic and abdominal dimensions to correctly interpret the
changes in pressure.
Maximum relaxation rate. The rate at which Pdi returns to
baseline with each spontaneous breath, or after stimulation or
after a sniff maneuver, can be used as a predictor of diaphragmatic fatigue (150, 151). Maximum relaxation rate has been
used in the ICU to study patients being weaned from mechanical ventilation (152). Weaning failure was associated with a
slowing of the maximum relaxation rate of the respiratory
muscles (Pes) and of the diaphragm, indirectly suggesting that
the applied load in these patients exceeded the capacity of the
respiratory muscles (see RELAXATION RATE in Section 5 of this
Statement).
Phrenic nerve stimulation. Performance of a full forcefrequency curve for the detection of respiratory muscle fatigue is
Figure 6. Recordings illustrating the method used to assess patient

ventilator interaction. From top to bottom: flow; volume; airway pressure; pleural pressure (Ppl); and abdominal (gastric) pressure. Ineffective inspiratory efforts are revealed by negative swings in Ppl during
expiration (arrows) that are not matched with a ventilator-assisted
breath. The concomitant declines in Ppl and abdominal pressure at the
beginning of expiration suggest that Ppl was not dominated by abdominal muscle contraction. PSV 15 and PSV 20 indicate pressure support ventilation of 15 and 20 cm H2O, respectively. Reprinted by permission from Reference 106.
impractical in critically ill patients (153). Measurement of the

response to a single supramaximal stimulation of the phrenic
nerve, i.e., twitch stimulation, is an appealing method of monitoring changes in diaphragmatic contractility (153). This approach has the considerable attraction of being independent
of volitional influences (154). However, systematic investigations of twitch stimulation of the phrenic nerves in critically ill
patients have yet to be published. The recent introduction of
magnetic stimulation (155157), which has several advantages
over electrical stimulation of the phrenic nerves, should further increase the feasibility of twitch stimulation in the ICU
setting (see PHRENIC NERVE STIMULATION in Section 2 of this
Statement).
Diaphragmatic Electromyography
In mechanically ventilated patients, direct recording of the electrical activity of the crural diaphragm via an esophageal electrode has been used to quantify the activity of this muscle, and
to examine the precise timing of activation (54, 55, 158161)
(see RECORDING ELECTRODES in Section 3 of this Statement).
PREDICTION OF WEANING
Management of the difficult-to-wean patient is one of the most

challenging problems in critical care medicine (2, 3). Overall,
approximately 40% of the time that a patient spends receiving
mechanical ventilation is devoted to the weaning process, and
in certain disease states, such as COPD, the weaning process
accounts for approximately 60% of ventilator time (162). Deciding the best time to initiate weaning is especially important.
If weaning onset is postponed, the patient is placed at an increased risk of life-threatening, ventilator-induced complications. If weaning is commenced prematurely, the patient may
617
suffer severe cardiopulmonary or psychological decompensation, which sets the patient back in his or her clinical course.
Although careful clinical assessment is necessary in deciding
when to discontinue mechanical ventilation, this alone is not
sufficient, as demonstrated by Stroetz and Hubmayr (163).
They studied 31 patients being weaned by gradual reductions
in the level of pressure support. The physician in charge of
each patient was asked to predict a patients ability to sustain
unassisted breathing without distress for 1 hour. Of 22 patients
whom the physicians thought likely to fail a weaning trial, 11
were successfully weaned; of nine patients thought likely to be
successfully weaned, three failed the trial. This study provides
clear evidence of the need to employ objective tests, i.e., predictive indices, when deciding if a patient can tolerate the discontinuation of mechanical ventilation.
Several tests have been described to guide the timing and
pace of the weaning process. Because an imbalance between
respiratory mechanical load and respiratory muscle capability
appears to be the major cause of weaning failure (18, 19),
many of the weaning indices directly or indirectly relate to respiratory muscle function. Before assessing respiratory muscle
function in a ventilator-supported patient, it is imperative to
show that the patient is not likely to develop hazardous hypoxemia. Weaning should not be contemplated in a patient
with a PaO2 of less than 55 mm Hg at an inspired oxygen fraction (FIO2 ) of 0.40 or higher, because weaning failure is very
likely in such patients. However, many patients with satisfactory oxygenation fail weaning attempts, and, except in the patient with marked hypoxemia, indices of oxygenation, such as
the PaO2 /FIO2 ratio, are unreliable predictors of weaning outcome (164, 165).
Maximum inspiratory pressure, a measurement of inspiratory muscle strength (see VOLITIONAL TESTS OF RESPIRATORY
MUSCLE STRENGTH in Section 2 of this Statement), is a standard
weaning index. In an early study, Sahn and Lakshminarayan
(61) found that all patients with a PI,max value more negative
than 30 cm H2O were successfully weaned while all patients
with a PI,max less negative than 20 cm H2O failed a weaning
trial. Subsequent investigators have not found PI,max to provide such clear separation between weaning success and weaning failure patients (13, 18, 106, 164169). Studies with accessible data on the accuracy of PI,max are summarized in Table 1.
Because the studies differ in design (prospective, retrospective), method of performing PI,max (best of three attempts,
sustained occlusion for 20 seconds), method of weaning (trials
of spontaneous breathing, intermittent mandatory ventilation,
pressure support), and definitions of weaning success and failure, it is not surprising that accuracy of PI,max varied considerably among the studies. Nevertheless, it is possible to arrive
at some general conclusions. Sensitivity was approximately
0.80, meaning that approximately 80% of patients who succeeded in a weaning trial had a PI,max value that predicted
success (i.e., more negative than 30 cm H2O). However,
specificity was approximately 0.25, meaning that only a minority (25%) of patients who failed a weaning trial had a PI,max
that predicted weaning failure (i.e., less negative than 30 cm
H2O). Moreover, the ability to predict outcome was not improved by employing a standardized method of measuring
PI,max (18, 106, 165). Based on these data, PI,max measurements appear to be more helpful in understanding the reason
why a particular patient failed a weaning trial rather than in
deciding whether to attempt a weaning trial (see additional
discussion of PI,max in this section).
A minute ventilation of less than 10 L/minute, indicating
acceptable ventilatory requirements, is another standard weaning index (61). However, in most studies, minute ventilation
618
has been shown to be little better than chance in predicting

outcome (165, 166). More helpful is the partitioning of minute
ventilation into its VT and fR components, so as to formulate
the ratio of fR to VT, which can be used to quantitate the extent of rapid shallow breathing. The fR/VT ratio, measured
with a simple handheld spirometer attached to the endotracheal tube while the patient spontaneously breathes room air
for 1 minute, has been proposed as a reliable method of predicting weaning outcome (165). In an initial training data set
in 36 patients, an fR/VT ratio of 105 breaths/minute/L provided
the best separation between patients who were successfully
weaned and those in whom weaning failed. This threshold
value was then prospectively evaluated in 64 patients who
constituted the validation data set. The positive and negative predictive values were 0.78 and 0.95, respectively (165).
Analyzing the data with receiver-operating-characteristic
(ROC) curves, the area under the curve for the fR/VT ratio
(0.89) was the highest of 10 weaning indices evaluated in the
study. Of note, the area under the ROC curves of conventional weaning indices such as minute ventilation (0.40), the
PaO2/PAO2 ratio (0.48), and PI,max (0.61) were not significantly
greater than that of an arbitrary test that is expected a priori
to have no discriminating value (Figure 7). Jaechke and coworkers (170) reanalyzed the data of Yang and Tobin (165) in
terms of likelihood ratios (171). The likelihood ratio for an fR/
VT ratio of less than 80 breaths/minute/L is 7.53 (Table 2). In
other words, an fR/VT of less than 80 is 7.53 times more likely
to occur in a patient who is subsequently successfully weaned
than to occur in a patient who will fail a weaning trial; a likelihood ratio of 510 is considered to generate moderate and
usually useful shifts in pretest to posttest probability (171).
Conversely, an fR/VT ratio greater than 100 is only 0.04 as
likely to occur in a patient who will be successfully weaned as
in a patient who subsequently fails a weaning trial; a likelihood ratio of 0.10 is considered to generate large and often
conclusive changes in the probability of a given diagnosis (in
this case weaning failure).
In patients being weaned from mechanical ventilation, Sassoon and Mahutte (168) found that the fR/VT ratio had a sensitivity of 0.97 and a specificity of 0.40. The product of fR/VT
and airway occlusion pressure (P0.1) had the same sensitivity,
but specificity increased to 0.60. Disappointingly, the combination of fR/VT and P0.1 did not have a higher area under the
ROC curve than that achieved by fR/VT alone.
Epstein (172) studied 84 patients with an fR/VT ratio 100
breaths/minute/L and observed a positive-predictive value of
0.83. Fourteen patients required reintubation, and all but one
of these had developed a new problem such as upper airway
obstruction or congestive heart failure. Ten patients with an
fR/VT
100 were extubated, and four required reintubation
(negative predictive value 0.4).
VOL 166
2002
Chatila and coworkers (169) measured fR/VT with a handheld spirometer in 100 medical-cardiac patients being weaned
from mechanical ventilation. The fR/VT ratio was measured
during the first minute of spontaneous breathing, and the patients were then weaned using T-tube trials, CPAP, or pressure-support ventilation. After 3060 minutes, the fR/VT ratio
was again measured. The area under the ROC curve was
higher for the fR/VT ratio measured at 3060 minutes than
during the first minute of spontaneous breathing, 0.92 0.03
and 0.74 0.05, respectively. Interestingly, the area under the
ROC curve for 3060 minutes is similar to that reported by
Yang and Tobin (0.89) (165). This study by Chatila and coworkers (169) emphasizes the importance of not measuring
the fR/VT ratio during the first minute of spontaneous breathing, when respiratory drive may still be suppressed. It does not
necessarily mean that one needs to wait 3060 minutes to
reach a steady-state value that truly represents the patients
clinical status.
In a study of 218 extubated patients in a medical ICU, Epstein and Ciubotaru (173) showed that women have a higher
fR/VT ratio than men, which could not be explained by body
size, and that fR/VT was further increased in women with a
narrow endotracheal tube. Consequently, the false-negative
rate for fR/VT in women with a narrow endotracheal tube ( 7
mm internal diameter) was especially high. When comparing
data from different studies of weaning indices, the pretest
probability of a successful outcome becomes very important.
Epstein and Ciubotaru (173) deliberately selected patients with
a very high pretest probability of successful outcome (0.84),
because every patient tolerated a weaning trial and was extubated; in contrast, outcome was in considerable doubt (pretest
probability 0.56) in the patients studied by Yang and Tobin
(165), who were trying to identify the earliest point in time
that a patient could resume spontaneous ventilation. For identical false-negative (0.03) and true-negative rates (0.64) (166),
the posttest probability of a successful outcome for fR/VT
100 breaths/minute/L, calculated on the basis of Bayes theorem, is 20% for the patients of Epstein and Ciubotaru (173)
versus 5% for the patients of Yang and Tobin (165). In other
words, fR/VT is less helpful in cases when the physician
strongly suspects that the patient can tolerate weaning and extubation than when the physician is very doubtful about a patients outcome.
Recently, Ely and coworkers (129) investigated whether
the combination of predictive indices followed by a trial of
spontaneous breathing would hasten the pace of weaning.
Over a 9-month period, ventilator-supported patients in their
medical ICU and coronary care units were screened each
morning for five factors: PaO2/FIO2 ratio
200; PEEP 5 cm
H2O; fR/VT 105 breaths/minute/L; intact cough on suctioning; and absence of infusions of sedative or vasopressor agents.
TABLE 1. ACCURACY OF MAXIMAL INSPIRATORY PRESSURE AS A PREDICTOR OF WEANING OUTCOME

P I,max Threshold (cm H2O)
30
30
20
30
15
20
30
20
20
30
Sensitivity
Specificity
Positive Predictive Value
Negative Predictive Value
Reference Number
0.68
1.00
NA
NA
1.00
1.00
0.86
0.91
0.90
0.67
0
0
NA
NA
0.11
0.14
0.21
0.30
0.26
0.69
0.56
0.67
0.91
0.92
0.59
0.60
0.58
0.82
0.67
0.78
0
0
0.22
0.21
1.00
1.00
0.55
0.55
0.60
0.55
166
167
164
164
165
165
165
168
169
169
Definition of abbreviations: NA not available; PI,max maximum inspiratory pressure.
619
Figure 7. Receiver operating characteristic (ROC) curves for frequency/tidal volume (f/VT) ratio, CROP index (an index that integrates measurements of dynamic compliance, respiratory rate, arterial oxygenation, and maximal inspiratory pressure),
maximal
inspiratory pressure (PI,max), and minute ventilation ( VE) in 36 patients who were successfully weaned and 28 patients who failed a
weaning trial. The ROC curve is generated by plotting the proportion of true positive results against the proportion of false positive results for each value of a test. The curve for an arbitrary test that is expected a priori to have no discriminatory value appears as a diagonal
line, whereas a useful test has an ROC curve that rises rapidly and
reaches a plateau. The area under the curve (shaded) is expressed
(boxed) as a proportion of the total area. Reprinted by permission
from Reference 165.
Patients in the intervention group (n 149) who met all five

criteria underwent a 2-hour trial of spontaneous breathing
that same morning. If the patient did not develop clinical signs
of distress, employing objective criteria developed in earlier
studies of weaning techniques (127, 128), the trial was considered successful, and the patients physician was notified of this
result. The control group (n 151) had daily screening but
did not undergo the spontaneous breathing trial. Although patients in the intervention group had more severe disease, with
higher APACHE II and acute-lung-injury scores, their median duration of mechanical ventilation was 1.5 days less than
the control group (p 0.003), and they had lower rates of
complications (p 0.001) and reintubation (p 0.04) and
lower ICU charges (p 0.03). This study demonstrates that
the systematic use of weaning indices results in better patient
outcomes than reliance on the clinical judgment of the attending physician.
In the original description of the fR/VT ratio, the measurement was obtained with a hand-held spirometer while the patient was disconnected from the ventilator and breathed room
air for 1 minute (165). Some investigators have measured fR/
VT during partial ventilator assistance, such as pressure support. For example, Lee and coworkers (174) reported that
eight of nine patients who failed a weaning trial had an fR/VT
105 breaths/minute/L, and they concluded that fR/VT was unhelpful. Because pressure support is well known to decrease fR
and increase VT, it is not surprising that the fR/VT threshold
value developed during unassisted breathing would not apply
during pressure support ventilation.
Finally, the definition of the outcome event is very important in assessing the accuracy of a diagnostic test. Accumulating data suggest that the fR/VT ratio is more helpful in predicting the earliest point in time that a patient can breathe
spontaneously without assistance from a ventilator than in
predicting the need for reintubation after the patient has been
extubated (129, 172, 173, 175). That is, fR/VT appears to be
more helpful in deciding when to initiate the weaning process,
rather than deciding when to terminate it and extubate the patient.
CONCLUSION
This Section of the Statement has reviewed the tests of respiratory muscle function, direct and indirect, that are relevant to
the particular environment of the ICU. The scientific basis
and methodology of many of these tests are discussed in earlier Sections of the Statement. This article also focuses on the
crucial issue of weaning from mechanical ventilation.
Abnormalities of the pattern of breathing are common in
ICU patients, especially in those with respiratory muscle dysfunction. Tachypnea is a sensitive marker of deteriorating
clinical status, but is not specific. Paradoxical motion of the rib
cage and abdomen occurs with elevated respiratory load, but
is not diagnostic of respiratory muscle fatigue. Measurements
of maximum inspiratory pressure have poor reproducibility in
critically ill patients and are of limited use for decision making
in the ICU. Airway occlusion pressure (P0.1) is easy to measure in patients receiving assisted ventilation, and high values
signal increased respiratory motor output. The most direct
measures of patient effort are work of breathing and PTP, but
these measurements require considerable attention to detail
and their use is confined to research settings. A scooped contour of the airway pressure tracing during volume and flow
preset assisted ventilation indicates patient effort. The relative
TABLE 2. LIKELIHOOD OF SUCCESSFUL WEANING RELATED TO

fR/VT RATIO*
fR/VT Ratio
Likelihood Ratio
80
80100
100
Definition of abbreviations: fR respiratory frequency; VT tidal volume.
* Based on data of Yang and Tobin (171).
7.53
0.77
0.04
620
swings in esophageal and gastric pressures provides insight

into the recruitment of different respiratory muscle groups.
Respiratory muscle contractility can be assessed by recording
transdiaphragmatic pressure provides in response to phrenic
nerve stimulation, a complex and invasive technique. Intrinsic
PEEPi, common in critically ill patients, can be measured using the EEO pressure in ventilated patients and by an esophageal balloon catheter in patients breathing spontaneously;
with both techniques, however, the value can be falsely elevated because of expiratory muscle activity.
The ICU setting where tests of respiratory muscle function
have greatest application is in the prediction of weaning outcome. Traditional predictors, such as maximum inspiratory
pressure, vital capacity, and minute ventilation, are frequently
falsely positive and negative. Although patients failing a
weaning trial may have an elevated tensiontime index and
airway occlusion pressure, these tests have not gained popularity in everyday ICU practice. The ratio of respiratory frequency to tidal volume is the most reliable simple predictor of
weaning outcome.
References
1. Aldrich TK, Prezant DJ. Indications for mechanical ventilation. In: Tobin MJ, editor. Principles and practice of mechanical ventilation.
New York: McGraw-Hill; 1994. p. 155189.
2. Lessard MR, Brochard LJ. Weaning from ventilatory support. Clin
Chest Med 1996;17:475489.
3. Tobin MJ, Alex CG. Discontinuation of mechanical ventilation. In: Tobin MJ, editor. Principles and practice of mechanical ventilation.
4. Durbin CG, Kopel RF. A case-control study of patients readmitted to
the intensive care unit. Crit Care Med 1993;21:15471553.
5. Schein RMH, Hazday N, Pena M, Ruben BH, Sprung CL. Clinical antecedents to in hospital cardiopulmonary arrest. Chest 1990;98:1388
1392.
6. Gravelyn TR, Weg JR. Respiratory rate as an indicator of acute respiratory dysfunction. JAMA 1980;244:11231125.
7. McFadden JP, Price RC, Eastwood HD, Briggs RS. Raised respiratory
rate in elderly patients: a valuable physical sign. BMJ 1982;284:626
627.
8. ODonnell DE, Webb KA. Measurement of respired volume and flow.
In: Tobin MJ, editor. Principles and practice of intensive care monitoring. New York: McGraw-Hill; 1998. p. 6380.
9. Gilbert R, Auchincloss JH, Brodsky J, Boden W. Changes in tidal volume, frequency, and ventilation induced by their measurement. J
Appl Physiol 1972;33:252254.
10. Perez W, Tobin MJ. Separation of factors responsible for change in
breathing pattern induced by instrumentation. J Appl Physiol 1985;
59:15151520.
11. Tobin MJ, Chadha TS, Jenouri G, Gazeroglu HB, Sackner MA.
Breathing patterns. 1. Normal subjects. Chest 1983;84:202205.
12. Tobin MJ, Mador MJ, Guenther SM, Lodato RF, Sackner MA. Variability of resting respiratory drive and timing in healthy subjects. J
Appl Physiol 1988;65:309317.
13. Tobin MJ, Perez W, Guenther SM, Semmes BJ, Mador MJ, Allen SJ,
Lodato RF, Dantzker DR. The pattern of breathing during successful and unsuccessful trials of weaning from mechanical ventilation.
Am Rev Respir Dis 1986;134:11111118.
14. Cohen C, Zagelbaum G, Gross D, Roussos C, Macklem PT. Clinical manifestations of inspiratory muscle fatigue. Am J Med 1982;73:308316.
15. Gallagher CG, Im Hof V, Younes MK. Effect of inspiratory muscle fatigue on breathing pattern. J Appl Physiol 1985;59:11521158.
16. Yan S, Sliwinski P, Gauthier AP, Lichros I, Zakynthinos S, Macklem
PT. Effect of global inspiratory muscle fatigue on ventilatory and respiratory muscle responses to CO2. J Appl Physiol 1993;75:13711377.
17. Mador MJ, Tobin MJ. The effect of inspiratory muscle fatigue on
breathing pattern and ventilatory response to CO2. J Physiol 1992;
455:1732.
18. Jubran A, Tobin MJ. Pathophysiological basis of acute respiratory distress in patients who fail a trial of weaning from mechanical ventilation. Am J Respir Crit Care Med 1997;155:906915.
19. Vassilakopoulos T, Zakyntinos S, Roussos C. The tensiontime index
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
VOL 166
2002
and the frequencytidal volume ratio are the major pathophysiologic

determinants of weaning failure and success. Am J Respir Crit Care
Med 1988;158:378385.
Strickland JH, Hasson JH. A computer-controlled ventilator weaning
system. Chest 1991;100:10961099.
Strickland JH, Hasson JH. A computer-controlled ventilator weaning
system: a clinical trial. Chest 1993;103:12201226.
Dojat M, Harf A, Touchard D, LaForest M, Lemaire F, Brochard L.
Evaluation of a knowledge-based system providing ventilatory management and decision for extubation. Am J Respir Crit Care Med
1996;153:9971004.
Konno K, Mead J. Measurement of the separate volume changes of ribcage and abdomen during breathing. J Appl Physiol 1967;22:407422.
Ashutosh K, Gilbert R, Auchincloss JH, Peppi D. Asynchronous
breathing movements in patients with chronic obstructive pulmonary
disease. Chest 1975;67:553557.
Gilbert R, Ashutosh K, Auchincloss JH Jr, Rana S, Peppi D. Prospective study of controlled oxygen therapy. Chest 1977;71:456462.
Roussos C, Macklem PT. The respiratory muscles. N Engl J Med
1982;307:786797.
Grassino A, Macklem PT. Respiratory muscle fatigue and ventilatory
failure. Annu Rev Med 1984;35:625647.
Tobin MJ, Guenther SM, Perez W, Lodato RF, Mador MJ, Allen SJ,
Dantzker DR. KonnoMead analysis of ribcageabdominal motion
during successful and unsuccessful trials of weaning from mechanical
ventilation. Am Rev Respir Dis 1987;135:13201328.
Krieger BP, Ershowsky P. Noninvasive detection of respiratory failure
in the intensive care unit. Chest 1988;94:254261.
Tobin MJ, Perez W, Guenther SM, Lodato RF, Dantzker DR. Does
ribcage abdominal paradox signify respiratory muscle fatigue? J Appl
Physiol 1987;63:851860.
Clanton TL, Ameredes BT, Thomson DB, Julian MW. Sustainable inspiratory pressures over varying flows, volumes, and duty cycles. J
Appl Physiol 1990;69:18751882.
Wasson JH, Sox HC, Neff RK, Goldman L. Clinical prediction rules:
applications and methodological standards. N Engl J Med 1985;313:
793799.
Chevrolet JC, Deleamont P. Repeated vital capacity measurements as
predictive parameters for mechanical ventilation need and weaning
success in the GuillainBarr syndrome. Am Rev Respir Dis 1991;
144:814818.
Rieder P, Louis M, Jolliet P, Chevrolet J-C. The repeated measurement
of vital capacity is a poor indicator of the need for mechanical ventilation in myasthenia gravis. Intensive Care Med 1995;21:663668.
Rodarte JR, Rehder K. Dynamics of respiration. In: Macklem PT, Mead
J, editors. Handbook of physiology. Section 3: The respiratory system.
Bethesda, MD: American Physiological Society; 1986. p. 131144.
Tobin MJ. Mechanical ventilation. N Engl J Med 1994;330:10561061.
Luce JM. Reducing the use of mechanical ventilation. N Engl J Med
1996;335:19161917.
Slutsky AS. Mechanical ventilation. American College of Chest Physicians Consensus Conference. Chest 1993;104:18331859.
Bersten AD, Rutten AJ, Vedig AE, Skowronski GA. Additional work
of breathing imposed by endotracheal tube, breathing circuits and intensive care ventilators. Crit Care Med 1989;17:671677.
Shapiro M, Wilson RK, Casar G, Bloom K, Teague RB. Work of
breathing through different sized endotracheal tubes. Crit Care Med
1986;14:10281031.
Banner MJ, Downs JB, Kirby RR, Smith RA, Boysen PG, Lampotang
S. Effects of expiratory flow resistance on inspiratory work of breathing. Chest 1988;93:795799.
Banner MJ, Lampotang S, Boysen PG, Hurd TE, Desautels DA. Flow
resistance of expiratory positive-pressure valve systems. Chest 1986;
90:212217.
Katz JA, Zinn SE, Ozanne GM, Fairley HV. Pulmonary, chest wall,
and lungthorax elastances in acute respiratory failure. Chest 1981;
80:304311.
Pelosi P, Cereda M, Foti G, Giacomini M, Pesenti A. Alterations in
lung and chest wall mechanics in patients with acute lung injury: effects of positive end-expiratory pressure. Am J Respir Crit Care Med
1995;152:531537.
Coussa ML, Guerin C, Eissa NT, Corbeil C, Chasse M, Braidy J, Matar
N, Milic-Emili J. Partitioning of work of breathing in mechanically
ventilated COPD patients. J Appl Physiol 1993;75:17111719.
Armaganidis A, Roussos C. Measurement of the work of breathing in
the critically ill patient. In: Roussos C, editor. The thorax. Vol. 85.

47. Fleury B, Murciano D, Talamo D, Aubier C, Pariente R, Milic-Emili J.
Work of breathing in patients with chronic obstructive pulmonary
disease in acute respiratory failure. Am Rev Respir Dis 1985;132:822
827.
48. Marini JJ, Capps JS, Culver BH. The inspiratory work of breathing
during assisted mechanical ventilation. Chest 1985;87:613618.
49. Marini JJ, Rodriguez RM, Lamb V. Bedside estimation of the inspiratory work of breathing during mechanical ventilation. Chest 1986;89:
5663.
50. Marini JJ, Rodriguez RM, Lamb VJ. The inspiratory workload of patient-initiated mechanical ventilation. Am Rev Respir Dis 1986;134:
902909.
51. Marini JJ, Smith TC, Lamb VJ. External work output and force generation during synchronized intermittent mechanical ventilation. Am
Rev Respir Dis 1988;138:11691179.
52. Sassoon CSH, Light RW, Lodia R, Sieck GC, Mahutte CK. Pressure
time product during continuous positive airway pressure, pressure
support ventilation, and T-piece during weaning from mechanical
ventilation. Am Rev Respir Dis 1991;143:469475.
53. Ward ME, Corbeil C, Gibbons W, Newman S, Macklem PT. Optimization of respiratory muscle relaxation during mechanical ventilation.
Anesthesiology 1988;69:2935.
54. Brochard L, Pluswka F, Lemaire F. Improved efficacy of spontaneous
breathing with inspiratory pressure support. Am Rev Respir Dis 1987;
136:411415.
55. Brochard L, Harf A, Lorino H, Lemaire F. Inspiratory pressure support prevents diaphragmatic fatigue during weaning from mechanical ventilation. Am Rev Respir Dis 1989;139:513521.
56. Alberti A, Gallo F, Fongaro A, Valentia S, Rossi A. P0.1 is a useful parameter in setting the level of pressure support ventilation. Intensive
Care Med 1995;21:547553.
57. Mancebo J. Weaning from mechanical ventilation. Eur Respir J 1996;9:
19231931.
58. Milic-Emili J. Measurement of pressures in respiratory physiology. In:
Otis AB, editor. Techniques in respiratory physiology. P4/II. County
Clare, Ireland: Elsevier Scientific Publishers Ireland; 1984. p. 122.
59. Rossi A, Polese G, Milic-Emili J. Mechanical ventilation in the passive
patient. In: Derenne JP, Whitelaw WA, Similowski T, editors. Acute
respiratory failure in chronic obstructive pulmonary disease. Vol. 92.
60. Decramer M, Macklem PT. Pressure developed by the respiratory muscles. In: Roussos C, editor. The thorax. Vol. 85. New York: Marcel
Dekker; 1995. p. 10991126.
61. Sahn SA, Lakshminarayan S. Bedside criteria for discontinuation of
mechanical ventilation. Chest 1973;63:10021005.
62. Marini JJ, Smith TC, Lamb V. Estimation of inspiratory muscle strength
in mechanically ventilated patients: the measurement of maximal inspiratory pressure. J Crit Care 1986;1:3238.
63. Multz AS, Aldrich TK, Prezant DJ, Karpel JP, Hendler JM. Maximal
inspiratory pressure is not a reliable test of inspiratory muscle
strength in mechanically ventilated patients. Am Rev Respir Dis 1990;
142:529532.
64. Milic-Emili J, Whitelaw WA, Derenne JP. Occlusion pressure: a simple
measure of the respiratory centers output. N Engl J Med 1975;293:
10291030.
65. Whitelaw WA, Derenne JP. Airway occlusion pressure. J Appl Physiol
1993;74:14751483.
66. Derenne JP. P0.1: about the relevance of 100 milliseconds. Intensive
Care Med 1995;21:545546.
67. Marazzini L, Cavestri R, Gori D, Gatti L, Longhini E. Difference between mouth and esophageal occlusion pressure during CO2 rebreathing in chronic obstructive pulmonary disease. Am Rev Respir
Dis 1978;118:10271033.
68. Murciano D, Aubier M, Bussi S, Derenne JP, Pariente R, Milic-Emili J.
Comparison of esophageal, tracheal, and mouth occlusion pressure
in patients with chronic obstructive pulmonary disease during acute
respiratory failure. Am Rev Respir Dis 1982;126:837841.
69. Fernandez R, Blanch L, Artigas A. Respiratory center activity during
mechanical ventilation. J Crit Care 1991;6:102111.
70. Aubier M, Murciano D, Fournier M, Milic-Emili J, Pariente R, Derenne JP. Central respiratory drive in acute respiratory failure patients with chronic obstructive pulmonary disease. Am Rev Respir
Dis 1980;122:191199.
71. Murciano D, Boskowski J, Lecocguic Y, Milic-Emili J. Tracheal occlusion pressure: a simple index to monitor respiratory muscle fatigue
during acute respiratory failure in patients with chronic obstructive
pulmonary disease. Ann Intern Med 1988;108:800805.
621
72. Sassoon CSH, Te TT, Mahutte CK, Light RW. Airway occlusion pressure: an important indicator for successful weaning in patients with
chronic obstructive pulmonary disease. Am Rev Respir Dis 1987;
135:107113.
73. Sassoon CSH, Mahutte CK, Simmons DH, Light RW. Work of breathing and airway occlusion pressure during assist-mode mechanical
ventilation. Chest 1988;3:571576.
74. Conti G, Cinnella G, Barboni E, Lemaire F, Harf A, Brochard L. Estimation of occlusion pressure during assisted ventilation in patients
with intrinsic PEEP. Am J Respir Crit Care Med 1996;154:907912.
75. Kuhlen R, Hausmann S, Pappert D, Slama K, Rossaint R, Falke K. A
new method for P0.1 measurement using standard respiratory equipment. Intensive Care Med 1995;21:554560.
76. Tobin MJ. Monitoring respiratory mechanics in spontaneous breathing
patients. In: Tobin MJ, editor. Principles and practice of intensive
care monitoring. New York: McGraw-Hill; 1998. p. 617654.
77. Agostoni E, Mead J. Statics of the respiratory system. In: Fenn WO,
Rahn H, editors. Handbook of physiology. Section 3, Vol. I: Respiration. Washington, DC: American Physiological Society; 1964. p. 387
409.
78. Jardin F, Genevray B, Brun-Ney D, Bourdarias J-P. Influence of lung
and chest wall compliances on transmission of airway pressure to the
pleural space in critically ill patients. Chest 1985;88:653658.
79. Jubran A, Van de Graaff WB, Tobin MJ. Variability of patientventilator interaction with pressure support ventilation in patients with
chronic obstructive pulmonary disease. Am J Respir Crit Care Med
1995;152:129136.
80. Roussos C, Campbell EJM. Respiratory muscle energetics. In: Handbook of physiology. Section 3, Vol. 3: Respiration. Bethesda, MD:
American Physiological Society; 1986. p. 481509.
81. Fabry B, Guttmann J, Eberhard L, Bauer T, Haberthr C, Wolff G. An
analysis of desynchronization between the spontaneously breathing
patient and ventilator during inspiratory pressure support. Chest
1995;107:13871394.
82. Nava S, Bruschi C, Fracchia C, Braschi A, Rubini F. Patientventilator
interaction and inspiratory effort during pressure support ventilation
in patients with different pathologies. Eur Respir J 1997;10:177183.
83. Leung P, Jubran A, Tobin MJ. Comparison of assisted ventilator
modes on triggering, patient effort and dyspnea. Am J Respir Crit
Care Med 1997;155:19401948.
84. Barnard PA, Levine S. Critique on application of diaphragmatic time
tension index to spontaneously breathing humans. J Appl Physiol
1986;54:10671072.
85. Brochard L, Isabey D, Piquet J, Amaro P, Mancebo J, Messadi AA,
Brun-Buisson C, Rauss A, Lemaire F, Harf A. Reversal of acute exacerbations of chronic obstructive lung disease by inspiratory assistance with a face mask. N Engl J Med 1990;323:15231530.
86. Giuliani R, Mascia L, Recchia F, Caracciolo A, Fiore T, Ranieri VM.
Patientventilator interaction during synchronized intermittent mandatory ventilation. Am J Respir Crit Care Med 1995;151:19.
87. Ranieri VM, Mascia L, Petruzelli V, Bruno F, Brienza A, Giuliani R.
Inspiratory effort and measurement of dynamic intrinsic PEEP in
COPD patients: effects of ventilator triggering systems. Intensive
Care Med 1995;21:896903.
88. Ranieri V, Giuliani R, Mascia L, Grasso S, Petruzelli V, Puntillo N,
Perchiazzi G, Fiore T, Brienza A. Patientventilator interaction during acute hypercapnia: pressure support vs. proportional assist ventilation. J Appl Physiol 1996;81:426436.
89. Sassoon CSH, Lodia R, Rheeman CH, Kuei HJ, Light RN, Mahutte
CK. Inspiratory muscle work of breathing during flow-by, demandflow, and continuous-flow systems in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis 1992;145:12191222.
90. Sassoon CSH, Del Rosario N, Fei R, Rheeman CH, Gruer SE, Mahutte
CK. Influence of pressure- and flow-triggered synchronous intermittent mandatory ventilation on inspiratory muscle work. Crit Care
Med 1994;22:19331941.
91. Petrof BJ, Legar M, Goldberg P, Milic-Emili J, Gottfried SB. Continuous positive airway pressure reduces work of breathing and dyspnea during weaning from mechanical ventilation in severe chronic
obstructive pulmonary disease (COPD). Am Rev Respir Dis 1990;141:
281289.
92. Cinnella G, Conti G, Lofaso F, Lorino H, Harf A, Lemaire F, Brochard
L. Effects of assisted ventilation on the work of breathing: volumecontrolled versus pressure-controlled ventilation. Am J Respir Crit
Care Med 1996;153:10251033.
93. Aslanian P, El Atrous S, Isabey D, Valente E, Corsi D, Harf A, Lemaire F, Brochard L. Effects of flow triggering on breathing effort
622
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.

during partial ventilatory support. Am J Respir Crit Care Med 1998;
157:135143.
Appendini L, Patessio A, Zanaboni S, Carone M, Gukov B, Donner
CF, Rossi A. Physiologic effects of positive end-expiratory pressure
during mask ventilatory assistance in patients with acute exacerbation of COPD. Am J Respir Crit Care Med 1994;149:10691076.
Lessard MR, Lofaso F, Brochard L. Expiratory muscle activity increases intrinsic positive end-expiratory pressure independently of
dynamic hyperinflation in mechanically ventilated patients. Am J Respir
Crit Care Med 1995;151:562569.
Banner MJ, Blanch PB, Kirby RR. Imposed work of breathing and
methods of triggering a demand-flow, continuous positive airway pressure system. Crit Care Med 1993;21:183190.
Banner MJ, Kirby RR, Kirton OC, DeHaven CB, Blanch PB. Breathing frequency and pattern are poor predictors of work of breathing
in patients receiving pressure support ventilation. Chest 1995;108:
13381344.
Gluck EH, Barkoviak MJ, Balk RA, Casey LC, Silver MR, Bone RC.
Medical effectiveness of esophageal balloon pressure manometry in
weaning patients from mechanical ventilation. Crit Care Med 1995;23:
504509.
Rossi A, Polese G, Brandi G, Conti G. The intrinsic positive end expiratory pressure (PEEPi): physiology, implications, measurement, and
treatment. Intensive Care Med 1995;21:522536.
Ranieri VM, Grasso S, Fiore T, Giuliani R. Auto-positive end-expiratory pressure and dynamic hyperinflation. Clin Chest Med 1996;17:
379394.
Pepe PE, Marini JJ. Occult positive end-expiratory pressure in mechanically ventilated patients with airflow obstruction. Am Rev Respir
Dis 1982;126:166170.
Rossi A, Gottfried SB, Zocchi L, Higgs BD, Lennox S, Calverley PMA,
Begin P, Grassino A, Milic-Emili J. Measurement of static compliance of the total respiratory system in patients with acute respiratory
failure during mechanical ventilation: the effect of intrinsic PEEP.
Brandolese R, Broseghini C, Polese G, Bernasconi M, Brandi G, MilicEmili J, Rossi A. Effects of intrinsic PEEP on pulmonary gas exchange in mechanically ventilated patients. Eur Respir J 1993;6:358
363.
Kawagoe Y, Permutt S, Fessler HE. Hyperinflation with intrinsic PEEP
and respiratory muscle blood flow. J Appl Physiol 1994;77:2440
2448.
Broseghini C, Brandolese R, Poggi R, Polese G, Manzin E, Milic-Emili
J, Rossi A. Respiratory mechanics during the first day of mechanical
ventilation in patients with pulmonary edema and chronic airway obstruction. Am Rev Respir Dis 1988;138:355361.
Appendini L, Purro A, Patessio A, Zanaboni S, Carone M, Spada E,
Donner CF, Rossi A. Partitioning of inspiratory muscle workload
and pressure assistance in ventilator-dependent COPD patients. Am
Rossi A, Appendini L, Poggi R, Ganassini A, Brandolese R, Luzzani
A. Acute bronchial asthma: indications for intensive care. Eur Respir
Rev 1993;3:400403.
Rossi A, Appendini L. Wasted efforts and dyssynchrony: the patient
ventilator battle is back? Intensive Care Med 1995;21:867870.
Nava S, Bruschi C, Rubini F, Palo A, Iotti G, Braschi A. Respiratory
response and inspiratory effort during pressure support ventilation
in COPD patients. Intensive Care Med 1995;21:871879.
Gottfried SB. The role of PEEP or CPAP in the mechanically ventilated COPD patient. In: Roussos C, editor. The thorax. Vol. 85. New
York: Marcel Dekker; 1995. p. 24712500.
Rossi A, Polese G, Milic-Emili J. Monitoring respiratory mechanics in
ventilator-dependent patients. In: Tobin MJ, editor. Principles and
practice of intensive care monitoring. New York: McGraw-Hill;
1998. p. 553596.
Smith T, Marini JJ. Impact of PEEP on lung mechanics and work of
breathing in severe airflow obstruction. J Appl Physiol 1988;65:1488
1499.
Gottfried SB, Rossi A, Higgs BD, Calverley PMA, Zocchi L, Bozic C,
Milic-Emili J. Non-invasive determination of respiratory system mechanics during mechanical ventilation for acute respiratory failure.
Bernasconi M, Ploysongsang Y, Gottfried SB, Milic-Emili J, Rossi A.
Respiratory compliance and resistance in mechanically ventilated
patients with acute respiratory failure. Intensive Care Med 1988;
14:547553.
Broseghini C, Brandolese R, Poggi R, Bernasconi M, Manzin E, Rossi
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
VOL 166
2002
A. Respiratory resistance and intrinsic positive end-expiratory pressure (PEEP) in patients with the adult respiratory distress syndrome
(ARDS). Eur Respir J 1988;1:726731.
Poggi R, Brandolese R, Bernasconi M, Manzin E, Rossi A. Doxofylline
and respiratory mechanics. Chest 1989;96:772778.
Bernasconi M, Brandolese R, Poggi R, Manzin E, Rossi A. Doseresponse
and time course effects of inhaled fenoterol on respiratory mechanics and arterial oxygen tension in mechanically ventilated patients
with chronic airflow obstruction. Intensive Care Med 1990;16:108
114.
Grootendorst AF, Lugtigheid G, van den Weygert EJ. Error in ventilator measurement of intrinsic PEEP: cause and remedy. Respir Care
1993;38:348350.
Iotti G, Braschi A. Respiratory mechanics in chronic obstructive pulmonary disease. In: Vincent JL, editor. Intensive care medical update. Berlin: Springer-Verlag; 1990. p. 223230.
Gottfried SB, Reisman H, Ranieri MV. A simple method for the measurement of intrinsic positive end-expiratory pressure during controlled and assisted modes of mechanical ventilation. Crit Care Med
1992;20:621629.
Baydur A, Behrakis PK, Zin WA, Jaeger MJ, Milic-Emili J. A simple
method for assessing the validity of the esophageal balloon technique. Am Rev Respir Dis 1982;126:788791.
Maltais F, Reissmann H, Navalesi P, Hernandez P, Gursahaney A,
Ranieri VM, Sovilij M, Gottfried SB. Comparison of static and dynamic measurements of intrinsic PEEP in mechanically ventilated
patients. Am J Respir Crit Care Med 1994;150:13181324.
Ninane V, Yernault JC, De Troyer A. Intrinsic PEEP in patients with
chronic obstructive pulmonary disease. Am Rev Respir Dis 1993;148:
10371042.
Yan S, Kayser B, Tobiasz M, Sliwinski P. Comparison of static and dynamic intrinsic positive end-expiratory pressure using the Campbell
diagram. Am J Respir Crit Care Med 1996;154:938944.
Gillespie DJ. Comparison of intraesophageal balloon pressure measurements with nasogastric esophageal balloon system in volunteers.
Zakynthinos SG, Vassilakopoulos T, Roussos C. The load of inspiratory muscles in patients needing mechanical ventilation. Am J Respir
Crit Care Med 1995;152:12481255.
Brochard L, Rauss A, Benito S, Conti G, Mancebo J, Rekik N, Gasparetto A, Lemaire F. Comparison of three methods of gradual withdrawal from ventilatory support during weaning from mechanical
ventilation. Am J Respir Crit Care Med 1994;150:896903.
Esteban A, Frutos F, Tobin MJ, Alia I, Solsona JF, Valverdu I, Fernandez R, de la Cal MA, Benito S, Tomas R, et al. A comparison of four
methods of weaning patients from mechanical ventilation. N Engl J
Med 1995;332:345350.
Ely EW, Baker AM, Dunagan DP, Burke HL, Smith AC, Kelly PT,
Johnson MM, Browder RW, Bowton DL, Haponik EF. Effect of the
duration of mechanical ventilation of identifying patients capable of
breathing spontaneously. N Engl J Med 1996;335:18641869.
Mancebo J, Isabey D, Lorino H, Lofaso F, Lemaire F, Brochard L.
Comparative effects of pressure support ventilation and intermittent
positive pressure breathing (IPPB) in non intubated healthy subjects.
Eur Respir J 1995;8:19011909.
Mancebo J, Amaro P, Mollo JL, Lorino H, Lemaire F, Brochard L.
Comparison of the effects of pressure support ventilation delivered
by three different ventilators during weaning from mechanical ventilation. Intensive Care Med 1995;21:913919.
Mancebo J, Amaro P, Lorino H, Lemaire F, Harf A, Brochard L. Effects of albuterol inhalation on the work of breathing during weaning
from mechanical ventilation. Am Rev Respir Dis 1991;144:95100.
Polese G, Massara A, Poggi R, Brandolese R, Brandi G, Rossi A.
Flow-triggering reduces inspiratory effort during weaning from mechanical ventilation. Intensive Care Med 1995;21:682686.
Sassoon CSH, Giron AE, Ely EA, Light RW. Inspiratory work of
breathing on flow by and demand flow continuous positive airway
pressure. Crit Care Med 1989;17:11081114.
Sassoon CSH, Gruer SE. Characteristics of the ventilator pressure- and
flow-trigger variables. Intensive Care Med 1995;21:159168.
Banner MJ, Kirby RR, Gabrielli A, Blanch PB, Layon AJ. Partially
and totally unloading respiratory muscles based on real-time measurements of work of breathing. Chest 1994;106:18351842.
Petros AJ, Lamond CT, Bennett D. The Bicore pulmonary monitor.
Anaesthesia 1993;48:985988.
Banner MJ, Kirby RR, Blanch PB. Differentiating total work of
breathing into its component parts. Chest 1996;109:11411143.

139. Kimura T, Takezawa J, Nishiwaki K, Shimada Y. Determination of the
optimal pressure support level evaluated by measuring transdiaphragmatic pressure. Chest 1991;100:112117.
140. Roussos C, Macklem PT. Diaphragmatic fatigue in man. J Appl Physiol
1977;43:189197.
141. Pourriat JL, Lamberto C, Hoang P, Fournier JL, Vasseur B. Diaphragmatic fatigue and breathing pattern during weaning from mechanical
ventilation in COPD patients. Chest 1986;90:703707.
on human diaphragmatic fatigue. J Appl Physiol 1982;53:11901195.
143. Gilbert R, Auchincloss JH Jr. Relationship of rib cage and abdomen
motion to diaphragm function during quiet breathing. Chest 1981;80:
607612.
144. Diehl JL, Lofaso F, Deleuze P, Similowski T, Lemaire F, Brochard L.
Clinically relevant diaphragmatic dysfunction after cardiac operations. J Thorac Cardiovasc Surg 1994;107:487498.
145. Ford GT, Whitelaw W, Rosenal TW, Cruse PJ, Guenter C. Diaphragm
dysfunction after upper abdominal surgery in humans. Am Rev
Respir Dis 1983;127:431436.
146. Simonneau G, Vivien A, Sartene R, Kunstlinger F, Samii K, Noviant Y,
Duroux P. Diaphragm dysfunction induced by upper abdominal surgery: role of post-operative pain. Am Rev Respir Dis 1983;128:899903.
147. Dureuil B, Desmonts J-M, Mankikian B, Prokocimer C. The effects of
aminophylline on diaphragmatic dysfunction after upper abdominal
surgery. Anesthesiology 1985;62:242246.
148. Macklem PT. The assessment of diaphragmatic contractility. Anesthesiology 1985;62:229230.
149. Ninane V, Rypens F, Yernault JC, De Troyer A. Abdominal muscle
use during breathing in patients with chronic airflow obstruction. Am
Rev Respir Dis 1992;992:1621.
150. Esau SA, Bellemare F, Grassino A, Permutt S, Roussos C, Pardy R.
Appl Physiol 1983;54:13531360.
151. Esau SA, Bye PT, Pardy RL. Changes in rate of relaxation of sniffs with
152. Goldstone J, Green M, Moxham J. Maximum relaxation rate of the diaphragm during weaning from mechanical ventilation. Thorax 1994;
49:5460.
153. Tobin MJ, Laghi F. Monitoring of respiratory muscle function. In: Tobin MJ, editor. Principles and practice of intensive care monitoring.
154. Similowski T, Yan S, Gauthier AP, Macklem PT, Bellemare F. Contractile properties of the human diaphragm during chronic hyperinflation. N Engl J Med 1991;325:917923.
67:13111318.
156. Wragg S, Aquilina R, Moran J, Riding M, Hamnegard C, Fearn T,
Green M, Moxham J. Comparison of cervical magnetic stimulation
and bilateral percutaneous electrical stimulation of the phrenic nerves
in normal subjects. Eur Respir J 1994;7:17881792.
157. Laghi F, Harrison M, Tobin MJ. Comparison of magnetic and electrical
phrenic nerve stimulation in assessment of diaphragmatic contractility. J Appl Physiol 1996;80:17311742.
623
158. Flick GR, Bellamy PE, Simmons DH. Diaphragmatic contraction during assisted mechanical ventilation. Chest 1989;96:130135.
159. Annat GJ, Viale JP, Derymez CP, Bouffard YM, Delafosse BX, Motin
JP. Oxygen cost of breathing and diaphragmatic pressuretime index. Measurement in patients with COPD during weaning with pressure support ventilation. Chest 1990;98:411414.
160. Imsand C, Feihl F, Perret C, Fitting JW. Regulation of inspiratory neuromuscular output during synchronized intermittent mechanical ventilation. Anesthesiology 1994;80:1322.
161. Nava S, Ambrosino N, Rubini F, Fracchia C, Rampulla C, Torri G, Calderini E. Effect of nasal pressure support ventilation and external
PEEP on diaphragmatic activity in patients with severe stable COPD.
Chest 1993;103:143150.
162. Esteban A, Alia I, Ibanez J, Benito S, Tobin MJ, Spanish Lung Failure
Collaborative Group. Modes of mechanical ventilation and weaning:
a national survey of Spanish hospitals. Chest 1994;106:11881193.
163. Stroetz RW, Hubmayr R. Tidal volume maintenance during weaning with
pressure support. Am J Respir Crit Care Med 1995;152:10341040.
164. Krieger BP, Erschowsky PF, Becker DA, Hazeroglu HB. Evaluation of
conventional criteria for predicting successful weaning from mechanical
ventilatory support in elderly patients. Crit Care Med 1989;17:858861.
165. Yang K, Tobin MJ. A prospective study of indexes predicting the outcome of trials of weaning from mechanical ventilation. N Engl J Med
1991;324:14451450.
166. Tahvanainen J, Salenpera M, Nikki P. Extubation criteria after weaning from intermittent mandatory ventilation and continuous positive
airway pressure. Crit Care Med 1983;11:702707.
167. Fiastro JF, Habib MP, Shon BY, Campbell SC. Comparison of standard weaning parameters and mechanical work of breathing in mechanically ventilated patients. Chest 1988;94:232238.
168. Sassoon CSH, Mahutte CK. Airway occlusion pressure and breathing
pattern as predictors of weaning outcome. Am Rev Respir Dis 1993;
148:860866.
169. Chatila W, Jacob B, Guaglionone D, Manthous CA. The unassisted respiratory ratetidal volume ratio accurately predicts weaning outcome. Am J Med 1996;101:6167.
170. Jaechke RZ, Meade MO, Guyatt GH, Keenan SP, Cook DJ. How to
use diagnostic test articles in the ICU: diagnosing weanability using
f/VT. Crit Care Med 1997;25:514521.
171. Sackett DL, Haynes RB, Guyatt GH, Tugwell P. Clinical epidemiology: a basic science for clinical medicine, 2nd ed. Boston: Little,
Brown; 1991. p. 145148.
172. Epstein SK. Etiology of extubation failure and the predictive value of
the rapid shallow breathing index. Am J Respir Crit Care Med 1995;
152:545549.
173. Epstein SK, Ciubotaru RL. Influence of gender and endotracheal tube
size on preextubation breathing pattern. Am J Respir Crit Care Med
1996;154:16471652.
174. Lee KH, Hui KP, Chan TB, Tan WC, Lim TK. Rapid shallow breathing
(frequencytidal volume ratio) did not predict extubation outcome.
Chest 1994;105:540543.
175. Vallverdu I, Calaf N, Subirana C, Net A, Benito S, Mancebo J. Clinical
characteristics, respiratory functional parameters, and outcome of a
two-hour T-piece trial in patients weaning from mechanical ventilation. Am J Respir Crit Care Med 1998;158:18551862.
624
VOL 166
2002
THIS OFFICIAL STATEMENT WAS PREPARED BY AN AD-HOC SUBCOMMITTEE OF THE ASSEMBLY ON RESPIRATORY STRUCTURE AND
FUNCTION. THE SUBCOMMITTEE WAS CHAIRED BY ALEX GRASSINO (ATS CO-CHAIR) AND JOHN MOXHAM (ERS CO-CHAIR).
MEMBERS OF THE SUBCOMMITTEE ARE:
Alex E. Grassino, ATS Co-Chair, Centre Hospitalier de
lUniversit de Montreal and Meakins-Christie Laboratories, McGill University, Montreal, Quebec, Canada
John Moxham, ERS Co-Chair, Department of Respiratory
Medicine, Kings College Hospital, London, United Kingdom
THOMAS K. ALDRICH, Pulmonary Section, Department of
Medicine, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, New York
JULIAN ALLEN, Childrens Hospital of Philadelphia, Philadelphia, Pennsylvania
FRANOIS BELLEMARE, Department of Anesthesia, Centre
Hospitalier de lUniversit de Montreal, Montreal, Quebec, Canada
LAURENT BROCHARD, Hpital Henri Mondor, Creteil, France
PETER M. CALVERLY, Fazakerly Hospital, Liverpool, England
BARTOLOME R. CELLI, Pulmonary Critical Care, St. Elizabeths
Medical Center, Boston, Massachusetts
THOMAS CLANTON, Pulmonary Division, Ohio State University, Columbus, Ohio
NEIL J. DOUGLAS, Respiratory Medicine Unit, Department
of Medicine, Royal Infirmary, Edinburgh, United Kingdom
SANDRA ENGLAND, University of Medicine and Dentistry of
New Jersey-Robert Wood Johnson Medical School, New
Brunswick, New Jersey
MARC ESTENNE, Department of Chest Medicine, Erasme
University Hospital, Universit Libre de Bruxelles, Brussels, Belgium
JEAN WILL FITTING, Division de Pneumologie, Centre Hospitalier de lUniversit Vaudois, Lausanne, Switzerland
SIMON C. GANDEVIA, Prince of Wales Medical Research Institute, Sydney, Australia
CLAUDE GAULTIER, Department of Physiology, Hpital
Robert-Debre, Paris, France
G. JOHN GIBSON, Freeman Hospital, Newcastle upon Tyne,
United Kingdom
MALCOLM GREEN, Respiratory Muscle Laboratory, Royal

Brompton Hospital, London, United Kingdom
SAMUEL T. KUNA, Department of Medicine, University of
Pennsylvania and Pulmonary, Critical Care, and Sleep
Section, Philadelphia Veterans Administration Medical
Center, Philadelphia, Pennsylvania
STEPHEN H. LORING, Department of Anesthesia and Critical
Care, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts
DAVID K. MCKENZIE, Prince of Wales Medical Research Institute, Sydney, Australia
NEIL B. PRIDE, Thoracic Medicine, National Heart and Lung
Institute, Imperial College, London, United Kingdom
JEREMY ROAD, Respiratory Medicine, University Hospital,
Vancouver, British Columbia, Canada
JOSEPH R. RODARTE (deceased), Pulmonary Section, Department of Medicine, Baylor College of Medicine, Houston,
Texas
ANDREA ROSSI, Ospedale Riunti di Bergamo, Bergamo, Italy
NIKOLAOS SIAFAKAS, University Hospital, University of Crete,
Crete, Greece
GARY C. SIECK, Anesthesiology and Physiology, Mayo
Clinic, Rochester, Minnesota
THOMAS SIMILOWSKI, Laboratoire de Physio-Pathologie Respiratoire, Service de Pneumologie et Reanimation, Groupe
Hospitalier Piti Salptrire, Paris, France
CHRISTER SINDERBY, Department of Medicine, University of
Montreal, Quebec, Canada
GERALD S. SUPINSKI, University of Rochester Medical Center, Rochester, New York
MARTIN J. TOBIN, Pulmonary and Critical Care Division,
Hines Veterans Administration Hospital, Chicago, Illinois
ANDRE DE TROYER, Chest Service, Erasme University Hospital, Brussels, Belgium
WILLIAM WHITELAW, Department of Medicine, University of
Calgary, Alberta, Canada

ATS-ERS Statement On Respiratory Muscle Testing

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ATS-ERS Statement On Respiratory Muscle Testing

Uploaded by

Copyright:

Available Formats

American Thoracic Society/European Respiratory Society